RadioGraphics
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

DOI: 10.1148/rg.246045102
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Mahesh, M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Mahesh, M.
Related Collections
Right arrow Breast (Imaging and Interventional)
Right arrow Physics and Basic Science
RadioGraphics 2004;24:1747-1760
© RSNA, 2004

AAPM/RSNA Physics Tutorial

AAPM/RSNA Physics Tutorial for Residents

Digital Mammography: An Overview1

Mahadevappa Mahesh, MS, PhD

1 From The Russell H. Morgan Department of Radiology and Radiological Science, Johns Hopkins University School of Medicine, JHOC Suite 4235, 601 N Caroline St, Baltimore, MD 21287-0856. From the AAPM/RSNA Physics Tutorial at the 2003 RSNA scientific assembly. Received May 12, 2004; revision requested June 17 and received June 30; accepted July 15. The author has no financial relationships to disclose. Address correspondence to the author (e-mail: mmahesh@jhmi.edu).

Abstract

Recent advances in digital detector technology have paved the way to full-field digital mammography (FFDM) systems. The performance of these systems has evolved to the point where replacement of screen-film mammography (SFM) systems is becoming realistic. Despite some commonality between the two techniques, there are fundamental differences in how images are recorded, displayed, and stored. These differences necessitate an understanding of the principles of detection and the characteristics of digital images. Several approaches have been taken in the development of FFDM systems: (a) slot scanning with a scintillator and a charge-coupled device (CCD) array, (b) a flat-panel scintillator and an amorphous silicon diode array, (c) a flat-panel amorphous selenium array, (d) a tiled scintillator with fiberoptic tapers and a CCD array, and (e) photostimulable phosphor plates (computed radiography). Although the initial cost of an FFDM system is high compared with that of an SFM system, digital mammography has inherent advantages, such as wide dynamic range, reduction in recall rates, potential for reduction in radiation dose, increased patient throughput, postprocessing capability, and digital acquisition. These advantages and the rapidly occurring technologic developments will help establish FFDM as a mainstay of breast evaluation.

© RSNA, 2004

Index Terms: Breast neoplasms, radiography, 00.1215, 00.30 • Breast radiography, technology, 00.1215 • Physics • Radiography, digital, 00.1215

Introduction

Breast cancer is one of the most highly diagnosed cancers among women in the United States (1). More than one in eight women are diagnosed with breast cancer during their lifetime. Early diagnosis plays a critical role in reducing mortality and improving prognosis of this disease. The seriousness of invasive breast cancer increases with the size and extent of invasion at first diagnosis. Therefore, proper diagnosis is very critical for the treatment. High-quality images are essential.

In the past two to three decades, mammography has become the most sensitive technique for detecting nonpalpable lesions. Screening mammography has been shown to reduce breast cancer mortality by approximately 18%–30% in the past decade (2,3). The decline in the rate of breast cancer death in the past few years may be due in part to high-quality images obtained with screen-film mammography (SFM) systems. Although images from an SFM system are considered the standard of reference in diagnosing breast cancer, approximately 10%–20% of breast cancers detected at breast self-examination or physical examination are not visible at SFM. Also, only 5%–40% of the lesions detected with SFM and recommended for biopsy are found to be malignant (4–7). This indicates a high level of false-positives, resulting in unnecessary biopsies and related psychological stress to patients. Even though SFM is good, it is neither perfectly sensitive nor highly specific.

The mammographic process involves exposure of the breast to x-rays of mammographic energies (kilovolt peak) followed by the transmission and scattering of x-rays through breast tissue (Fig 1). The attenuated x-ray photons that pass through the grid interact with the image receptor and are finally absorbed as a latent image on the recording devices. After processing, the recorded images are then displayed for diagnosis and later archived (8). With SFM, the entire process is captured, displayed, and archived with a single medium, which is film. The widespread popularity of the SFM system is due to its many inherent advantages, such as high spatial resolution (up to 20 line pairs per millimeter), which can demonstrate fine spiculations and microcalcifications; high contrast, which allows visualization of subtle differences among soft-tissue densities (9,10); use of high-luminance view boxes, which improves visualization of dense tissues; and ease of display, rearrangement, and masking of film during diagnosis, which allows simultaneous display of images made during screening examinations and supplementary views of previous images on multiple panel illuminators (Table 1). Use of multiple image receptor sizes enables imaging breasts of different sizes. In addition, film acts as an efficient medium for long-term storage with low cost, and overall the advantages have established SFM as the standard of reference in mammography.



View larger version (32K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1.  Typical process of x-ray mammography.

 

View this table:
[in this window]
[in a new window]

  		TABLE 1. Advantages and Limitations of SFM

 
Despite these advantages, there are a number of inherent limitations with SFM. Among them are the limited dynamic range (Fig 2), trade-offs between dynamic range and contrast resolution, noise due to film granularity, and compromise between resolution and efficiency. Most important, film acts as the sole medium for acquisition, display, and storage of images. Suboptimal conditions in any one step can affect the overall image quality and limit the full capability of the mammography process. The limitation of SFM is further demonstrated with an image of a breast composed of a wide range of tissues (Fig 3) along with the mammography film characteristic response curve (Hurter and Driffield [H&D] curve) (11). As demonstrated in Figure 3, if the system is optimized for the dense part of the breast (the automatic exposure control sensor is positioned under the dense part of the breast), all other tissues fall on the upper plateau of the film response curve, making them impossible to visualize. In addition, technical factors such as film processing, developing, and image artifacts can limit the use of SFM. The inability to postprocess and optimize images often requires retakes, which can lead to multiple exposures, resulting in unnecessary radiation exposure to patients. Also, dense tissues with diffuse involvement of the breast with tumor tend to reduce the overall sensitivity of SFM. Digital mammography has the potential to overcome the limitations of SFM and has the potential to improve early breast cancer detection and lesion characterization (12–15).



View larger version (28K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2.  Typical response curves for SFM and digital mammography. SFM has a limited dynamic range, whereas digital mammography has a wider dynamic range.

 


View larger version (39K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3.  Limitations of SFM in imaging a breast composed of a wide range of tissues. Different regions of the breast image are represented according to the characteristic response of a typical mammographic film.

 
Digital mammography has been in the research and development phase for the past decade. The earliest experience with digital detectors comes from the stereotactic biopsy units (15). The advances in digital detectors offer improved detection due to the improved efficiency of absorption of incident x-ray photons. The linear response over a wide range of x-ray intensities and low system noise are some of the inherent advantages of digital detectors (Fig 2). Also, with digital mammography, the processes of image acquisition, displaying, and archiving are decoupled, unlike in SFM (Fig 4), thereby providing an opportunity to independently optimize each process. The other advantages of digital mammography include the wide dynamic range (1,000:1) compared with that of SFM (40:1), dynamic image manipulation, and the ability to postprocess, which allow improvement in lesion visibility. Soft-copy reading accompanied by computer-aided diagnosis (CAD) and three-dimensional imaging can improve the overall mammography process.



View larger version (19K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 4.  Unlike with SFM (left), each component of the mammographic process can be optimized with digital mammography (right).

 
The potential advantages of digital mammography over SFM are demonstrated in Figure 5, which shows a breast phantom imaged on an SFM system with automatic exposure control (AEC) settings, followed by two exposures at half and double the AEC technique. The phantom was then imaged on a full-field digital mammography (FFDM) system with technique factors identical to the SFM exposures. The window and levels were adjusted on the digital images prior to printing on a dry laser printer. Both the SFM and FFDM images are displayed on a high-luminance display system. The wide dynamic range and image processing capability enable visualization of the FFDM images with the three techniques, even though the image obtained at the lower exposure technique is noisier. On the other hand, the SFM images are either underexposed or overexposed with half and double the normal setting. This is due to the limited dynamic range of film and demonstrates the potential need for retakes. Figure 6 shows a set of analog and digital images of a woman with moderately radiodense breast tissue. Analog images (SFM images) show poor penetration in the dense region compared with digital images. Digital images have improved contrast, demonstrate enhanced depiction of a suspicious mass, and allow better visualization of peripheral tissue and the skin line.



View larger version (55K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 5a.  (a-c) Images of a breast phantom obtained with an SFM system by using automatic exposure control (b), using one-half of the milliampere-seconds value (a), and using double the milliampere-seconds value (c). The images are displayed on a high-luminance view box. (d-f) Corresponding images obtained with a full-field digital mammography (FFDM) system. The effect of underexposure on image quality is increasing noise, whereas overexposure decreases noise at the expense of longer exposure and higher breast dose. With the digital mammograms (d-f), underexposure and overexposure do not affect image contrast.

 


View larger version (78K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 5b.  (a-c) Images of a breast phantom obtained with an SFM system by using automatic exposure control (b), using one-half of the milliampere-seconds value (a), and using double the milliampere-seconds value (c). The images are displayed on a high-luminance view box. (d-f) Corresponding images obtained with a full-field digital mammography (FFDM) system. The effect of underexposure on image quality is increasing noise, whereas overexposure decreases noise at the expense of longer exposure and higher breast dose. With the digital mammograms (d-f), underexposure and overexposure do not affect image contrast.

 


View larger version (85K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 5c.  (a-c) Images of a breast phantom obtained with an SFM system by using automatic exposure control (b), using one-half of the milliampere-seconds value (a), and using double the milliampere-seconds value (c). The images are displayed on a high-luminance view box. (d-f) Corresponding images obtained with a full-field digital mammography (FFDM) system. The effect of underexposure on image quality is increasing noise, whereas overexposure decreases noise at the expense of longer exposure and higher breast dose. With the digital mammograms (d-f), underexposure and overexposure do not affect image contrast.

 


View larger version (112K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 5d.  (a-c) Images of a breast phantom obtained with an SFM system by using automatic exposure control (b), using one-half of the milliampere-seconds value (a), and using double the milliampere-seconds value (c). The images are displayed on a high-luminance view box. (d-f) Corresponding images obtained with a full-field digital mammography (FFDM) system. The effect of underexposure on image quality is increasing noise, whereas overexposure decreases noise at the expense of longer exposure and higher breast dose. With the digital mammograms (d-f), underexposure and overexposure do not affect image contrast.

 


View larger version (136K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 5e.  (a-c) Images of a breast phantom obtained with an SFM system by using automatic exposure control (b), using one-half of the milliampere-seconds value (a), and using double the milliampere-seconds value (c). The images are displayed on a high-luminance view box. (d-f) Corresponding images obtained with a full-field digital mammography (FFDM) system. The effect of underexposure on image quality is increasing noise, whereas overexposure decreases noise at the expense of longer exposure and higher breast dose. With the digital mammograms (d-f), underexposure and overexposure do not affect image contrast.

 


View larger version (128K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 5f.  (a-c) Images of a breast phantom obtained with an SFM system by using automatic exposure control (b), using one-half of the milliampere-seconds value (a), and using double the milliampere-seconds value (c). The images are displayed on a high-luminance view box. (d-f) Corresponding images obtained with a full-field digital mammography (FFDM) system. The effect of underexposure on image quality is increasing noise, whereas overexposure decreases noise at the expense of longer exposure and higher breast dose. With the digital mammograms (d-f), underexposure and overexposure do not affect image contrast.

 


View larger version (86K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 6a.  Analog and digital mammograms of a moderately dense breast. (a) Analog image demonstrates poor penetration in the dense region. (b) Digital image has improved contrast, shows a suspicious mass more clearly, and allows better visualization of peripheral tissue and the skin line. (Courtesy of Laurie Fajardo, MD, University of Iowa, Iowa City.)

 


View larger version (95K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 6b.  Analog and digital mammograms of a moderately dense breast. (a) Analog image demonstrates poor penetration in the dense region. (b) Digital image has improved contrast, shows a suspicious mass more clearly, and allows better visualization of peripheral tissue and the skin line. (Courtesy of Laurie Fajardo, MD, University of Iowa, Iowa City.)

 
The article intends to provide an overview of digital mammography including characteristics of various FFDM systems. Other topics discussed are storage and display of digital images, the economics of FFDM, regulation and accreditation of FFDM systems, clinical trials, phantom studies, challenges, and future advances. This article is the first in a series of three digital mammography topics, to be followed by articles on fundamentals of digital image quality and digital mammography displays.

Digital Mammography

Various approaches have been taken in the development of FFDM systems (13,14,16,17). Among them, one can classify systems as either indirect capture or direct capture. Indirect capture uses a two-step process whereby a scintillator such as cesium iodide (CsI) absorbs the x-rays and generates a light scintillation (similar to SFM), which is then detected by an array of photodiodes or charge-coupled devices (CCDs). Even though CsI crystals have a needlelike tubular structure and help channel light to the surface of the optical sensor, nevertheless, owing to imperfections in the needle structure, there is still some degree of light spread, which results in resolution degradation. For thin CsI scintillators, the absorption quantum efficiency is further diminished. With the direct capture process, the x-ray photons are directly captured by a photoconductor such as amorphous selenium (a-Se), which converts absorbed x-rays directly to a digital signal (18). The possibilities of resolution degradation due to light spread that are inherent with indirect capture are eliminated in these systems. In addition, spatial resolution with direct capture is limited to the pixel size and not to the thickness of the photoconductor.

The various approaches currently taken in the development of FFDM systems are (a) slot scanning with scintillators and CCD arrays, (b) a single flat-panel scintillator and an amorphous silicon (a-Si) diode array, (c) a flat-panel a-Se array, (d) tiled scintillators with fiberoptic tapers and mosaic CCD arrays, and (e) photostimulable phosphor plates (computed radiography).

Slot Scanning with a Scintillator and a CCD Array
In the first approach to FFDM, a narrow slot-detector and a narrow fan beam of x-rays are scanned synchronously across the full field of view to cover the entire breast (19). The commercial product based on this principle (SenoScan; Fischer Imaging, Denver, Colo) consists of two main subassemblies: the scanning gantry and the enclosure for the x-ray power supply and other electrical components (Fig 7). The system consists of phosphor (thallium-activated CsI) with a fiberoptic coupling to a CCD (20). The x-rays are collimated into a fan beam, matching the format of the detector array. Images are acquired by the detector scanning laterally across the breast in synchrony with the x-ray beam (Fig 7). The detector is only 1 cm wide in the scanning direction and consists of four CCDs abutted together, but is sufficiently long (22 cm) to include the entire breast in the anteroposterior direction (20).



View larger version (50K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 7a.  Full-size (a) and detail (b) photographs of the SenoScan system. The breastplate housing the detector assembly is curved to allow slot-scanning motion. This system is thus different from other mammography systems. (Courtesy of Fischer Imaging.)

 


View larger version (110K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 7b.  Full-size (a) and detail (b) photographs of the SenoScan system. The breastplate housing the detector assembly is curved to allow slot-scanning motion. This system is thus different from other mammography systems. (Courtesy of Fischer Imaging.)

 
The CCD matrix consists of 400 x 2,048 pixels and with a field of view of 21 x 29 cm produces images that are 4,096 x 5,625 pixels. The total scan time for the 30-cm image width is less than 6 seconds with an effective exposure time of approximately 1/30th of the scan time or 200 msec; consequently, the effect of patient motion on image sharpness is in most cases less noticeable than with the conventional systems. The detector element size is approximately 54 µm in standard-resolution mode and can be as low as 27 µm in high-resolution mode. As the slot geometry is inefficient in using the x-ray flux, the slot-scanning system uses a completely dedicated x-ray unit with a tungsten-rhenium anode tube and a choice of three filter materials: molybdenum, rhodium, or aluminum. Typically, a higher kilovoltage (31 kVp) is employed compared to other molybdenum or rhodium target systems. This system has been approved for clinical use in the United States for the past 3 years.

The slot-scan detector has a distinct advantage over the area detectors in that it is very compact and therefore the detector assembly costs less. Also, the system has excellent scatter rejection due to small breast volume exposed at any time. The need for a grid does not arise in these systems, thereby reducing the overall breast dose. However, these systems need powerful x-ray tubes and generators and elaborate signal readout and image reconstruction. Also, relative to other mammography systems, the system requires longer breast compression.

Flat-Panel Scintillator with an a-Si Diode Array
A second approach to FFDM is the use of a-Si thin-film transistor (TFT) arrays (21,22). In this system, a-Si diode arrays are constructed from a matrix of a-Si TFTs deposited on a glass substrate (Fig 8). CsI crystals are deposited as linear columns on the a-Si detector array, enabling minimal light divergence from the scintillator. Each light-sensitive diode element is connected by TFTs to a control and a data line, so that charge produced in a diode in response to light emission from the phosphor is read out and digitized.



View larger version (42K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 8.  Diagram of an a-Si TFT array. The digital detector array is constructed from an a-Si TFT matrix deposited on a glass substrate. The CsI scintillator is deposited on the a-Si detector, and each light-sensitive diode element is connected by TFTs to control and data lines, so that charge produced in the diode in response to light emission from the scintillator is read out and digitized. (Courtesy of GE Healthcare, Waukesha, Wis.)

 
The a-Si flat-panel system (Senographe 2000D; GE Healthcare) was the first FFDM system approved by the Food and Drug Administration (FDA) in the United States (Fig 9). It consists of 1,920 x 2,304 detector elements on a 19.2 x 23-cm area, with each pixel 100 µm, the largest pixel size of the available FFDM systems. The close bonding between the CsI(Tl) scintillator and a-Si photodetector ensures minimal light loss. Strong signal from the silicon diode array yields a higher detective quantum efficiency (DQE). The digital detector is linear over a wide range (105) of exposures. The compactness and simplicity of the detector design are some of the other salient features.



View larger version (144K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 9.  A clinical installation of the Senographe 2000D system. The breastplate housing the detector assembly is flat, similar to that of other mammography systems. In the background is the technologist’s workstation, where images are acquired and previewed before being transmitted to a review workstation located elsewhere. (Courtesy of GE Healthcare.)

 
The major limitations are the smaller image receptor size (difficulty in accommodating all breast sizes in a single exposure or the need for multiple exposures of large breasts) and the large pixel size (100 µm). However, there seems to be debate whether smaller pixel sizes (<100 µm) are essential for improving detection of calcifications. Smaller pixel sizes improve limiting spatial resolution but at the price of increased image noise and decreased signal-to-noise ratio for the same breast dose. It remains to be seen whether a digital system will be judged solely on the limiting spatial resolution or on the trade-offs between limiting spatial resolution and signal-to-noise ratio.

Flat-Panel a-Se Array
The only available direct digital detector for FFDM is that in the Selenia system (manufactured by Lorad, a division of Hologic, Bedford, Mass). The a-Se digital detector directly converts x-rays passing through the breast to electronic signals (18,23,24). The electron-hole pairs created in the process drift toward the respective polarized electrodes and are collected by pixel capacitors. a-Se is a good photoconductor with high x-ray absorption capability (>95%) in the mammographic energy ranges. The quantum efficiency is significantly higher than that of standard screens (50%–70%) and CsI scintillators (50%–80%). The a-Se is deposited directly onto the TFT substrate (Fig 10). The advantage of a-Se is that the response function maintains its sharpness (narrow line-spread function) even with increasing thickness.



View larger version (50K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 10.  Photograph of the a-Se digital detector used in the Selenia system. a-Se, a good photoconductor, is deposited directly onto the a-Si TFT substrate, enabling direct capture. (Courtesy of Hologic.)

 
The Selenia system has a detector with an approximately 25 x 29-cm field of view, the largest among commercially available systems, and a pixel size of 70 µm. Among its limitations are the demands for large storage space per examination, since overall spatial resolution depends on the large number of small pixels, the pixel size of the readout array, and the type of digital sampling scheme. The large digital detector is ideal for accommodating all breast sizes and presents no difficulty in positioning small breasts during imaging. The automatic exposure control system looks at the entire image area, then uses the area under the densest portion of the breast to determine the correct exposure (similar to an a-Si diode array system). The large image can be cropped to reduce data transfer time, display requirements, and storage space.

Tiled Scintillators with Fiberoptic Tapers and a CCD Array
Another system is based on modules that are similar to the digital detectors used in some stereotactic biopsy imaging systems (15). The detector module consists of a phosphor screen (CsI[Tl]), a CCD camera, and a fiberoptic taper to couple the light from the screen to the camera (Fig 11a). The CCD camera captures the light produced by the phosphor and converts it to an electrical signal. To image the entire breast, the system uses a 3 x 4 mosaic of these detector modules (12,14), each butted tightly to its neighbors (Fig 11b). The CCD array system (Lorad-Hologic) is 19 x 25 cm with a 40-µm pixel size. Because of the small pixel size and large field size, it also has the largest image matrix (4,500 x 6,000 pixels). The small pixel size results in high spatial resolution with the penalties of a large image format and the need to stitch interfaces between adjacent fiberoptic tapers, requiring additional image processing and possibly introducing small dead spaces within the active area. Although this system was approved for clinical use in the United States, it is no longer commercially available, as it was phased out in favor of the a-Se–based detector system acquired by the same parent company.



View larger version (48K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 11a.  (a) Diagram of the digital receptors in a small-field-of-view stereotactic biopsy unit. The scintillator is linked to a CCD by a minifying fiberoptic taper. (b) Photograph of the tiled CCD detector assembly (3 x 4 CCD array) used in the CCD array FFDM system. (Courtesy of Laurie Fajardo, MD.)

 


View larger version (60K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 11b.  (a) Diagram of the digital receptors in a small-field-of-view stereotactic biopsy unit. The scintillator is linked to a CCD by a minifying fiberoptic taper. (b) Photograph of the tiled CCD detector assembly (3 x 4 CCD array) used in the CCD array FFDM system. (Courtesy of Laurie Fajardo, MD.)

 
Photostimulable Phosphor Plates (Computed Radiography)
Another system is based on the original computed radiography (CR) product (Fuji Medical Systems, Stamford, Conn) introduced in 1981 with subsequent advances in imaging plate technology and image processing (25,26). The detector is a flexible plastic sheet coated with a photostimulable x-ray–absorbing phosphor material, typically barium fluorobromide (BaFBr:Eu2+). The imaging plates, available in standard mammographic cassette sizes, are located in cassettes for exposure in standard screen-film Bucky trays. In response to absorption of x-rays, electronic charges are stored proportionally in "traps" in the material of the phosphor, where they remain stable for some time. After exposure, the image is read by precision scanning of the imaging plate by a laser beam. The red laser light discharges the traps, causing stimulated emission of blue light. The blue light is collected by an efficient light guide and detected by a photomultiplier tube. The resulting signal is logarithmically amplified, digitized, and processed for film or soft-copy display. The imaging plate is erased by exposure to white light in the image reader for reuse.

The advantage of the CR system is that CR cassettes can be used to replace screen-film cassettes in existing mammography systems, without the need to replace the entire unit. Also, CR plates and readers can accommodate either 18 x 24-cm or 24 x 30-cm image receptor sizes, similarly to screen-film cassettes (unlike the single fixed size of other digital image receptors). However, the current implementation of high-resolution CR employs sampling on the order of 100 µm, yielding 100-µm pixel sizes in CR images. Also, the effective pixel sizes are influenced by the characteristics of the phosphor plate such as plate thickness, light diffusion within the plate, laser light scatter, and the diameter of the laser beam. New technology with a sharper readout laser and dual-side reader is expected to yield a 50-µm pixel size. Currently, the CR system is not approved for mammography in the United States. However, the system was used as an investigational device in a large clinical trial (the Digital Mammographic Imaging Screening Trial) that concluded recently.

Storage of Digital Images

FFDM systems present specific technical challenges regarding the image size and the storage and archiving of digital images by computer systems. A digital image is a two-dimensional grid of picture elements (pixels), which is defined by its size and bit depth. The size of an image is its length in pixels multiplied by its width in pixels, and the bit depth is the number of shades of gray that can be displayed. Signals are digitized or quantized into one of 2n intensity levels within each pixel, where n is the number of bits of digitization and is typically 12 or 14, depending on the system’s design. The demand for high spatial resolution in mammography requires smaller pixel size and higher digitization. Therefore, FFDM image size depends on the digital detector size, the number of pixels per image, and the type of digitization.

If the detected digital signals are digitized to 12 bits, it implies 212 or 4,096 signal values stored per pixel, and 14-bit digitization yields 214 or 16,384 signal values stored per pixel. Whether the system provides 12- or 14-bit digitization (8 bits = 1 byte), a digital detector of N pixels requires 2N bytes of storage (2 bytes per pixel). Image sizes and storage space required for various FFDM systems are shown in Table 2. In general, a typical screening mammography examination (two craniocaudal and two mediolateral oblique images) requires anywhere from 33 to more than 200 MB of computer storage space. Pixel size, field of view, maximum detection area, and other digital detector characteristics are listed in Table 3.


View this table:
[in this window]
[in a new window]

  		TABLE 2. Typical Image Sizes and Storage Requirements for FFDM Systems

 

View this table:
[in this window]
[in a new window]

  		TABLE 3. Scintillator, Detector Type, Pixel Size, Field of View, and Detector Area for FFDM Systems

 
Display of Digital Images

Various approaches are currently taken in displaying digital images. In the absence of soft-copy reading, digital images were printed on dry laser printers and read on high-luminance view boxes, similarly to SFM images. This approach does not allow the user to benefit from the full range of advantages available on digital displays. However, with the approval of soft-copy display, high-resolution (5-megapixel) computer monitors are commonly used.

The soft-copy display allows the flexibility of online image processing such as windowing and leveling. On some systems, readers can switch back and forth between preset display settings (such as "premium view" on the Senographe 2000D) to view digital breast images. The inherent advantages of soft-copy display include features such as magnification of certain areas by "thumbnailing" through the suspicious regions without additional exposure to the patient (Fig 12). The image processing capabilities on workstations, such as edge enhancement, postprocessing, and many other features, enhance clinical evaluation of digital mammograms. With increasing experience on display workstations, improved assessment and diagnosis without additional imaging and radiation exposure to patients can be attained.



View larger version (122K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 12.  Soft-copy display of a digital mammogram shows magnification of a suspicious area, which is evaluated without additional radiation exposure to the patient.

 
Most mammography centers still operate with SFM systems; during the past 2–3 years, many have been installing FFDM systems along with SFM systems, and only a few are in transition toward all-digital centers. During this transition time, the displays of both digital and conventional images are a challenge. With digital displays, in the absence of high-luminance displays, it is important to set up computer display monitors in areas with minimal ambient light (<5 lux). Therefore, in centers that have both FFDM and SFM systems, many designs for reading rooms are currently in practice. Figure 13 shows one such setup, where high-resolution computer monitors are placed at right angles to high-luminescence view boxes to minimize spectral reflections from each reading system to the other.



View larger version (114K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 13.  Typical layout of a mammography reading room in a center with both FFDM and SFM systems. To accommodate both digital and analog images, high-resolution computer monitors are placed at right angles to high-luminance view boxes to minimize spectral reflections from either of the reading systems. With digital displays, it is important to set up computer monitors in areas with minimal ambient light (<5 lux).

 
Economics of FFDM

As of now, rapid influx of digital mammography has not occurred mainly due to economics. Many centers may be waiting for the results of large clinical trials such as the Digital Mammographic Imaging Screening Trial (discussed in the Clinical Trials section). Although SFM systems cost well under $100,000, FFDM systems currently cost in the range of $300,000–$450,000. One attractive reason for centers to "go digital" is the higher reimbursement rates approved by Medicare in 2003 (27). Accordingly, the reimbursement rate for a typical screening mammography study is $132.06 when performed on an FFDM system, nearly 50% higher than when performed on a conventional SFM system ($82.77).

The costs of FFDM systems should be compared along with the inherent benefits of the digital technology prior to the purchase. Even though the accuracy of digital mammography reading in detecting cancer is yet to be determined (results from the recently concluded Digital Mammographic Imaging Screening Trial are expected soon), one should weigh the expected benefits of FFDM systems such as reduced recall rates, increased patient throughput, increased early detection of breast cancer, and decreased false-negative biopsy results against the cost of the system prior to purchase. Several factors can help justify the additional cost of FFDM, including increasing the caseload of each mammography room and decreasing film and processing costs.

Regulation and Accreditation of FFDM Systems

Mammography is one of the most highly regulated imaging modalities in the United States. Against this backdrop, FFDM systems underwent strict federal oversight by the Food and Drug Administration (FDA) (Rockville, Md) prior to approval for clinical use. Despite the existence of digital technology for quite some time, it was only in the early part of the year 2000 that the first FFDM system received approval for routine screening mammography (initially only hard-copy reading; soft-copy reading was approved in the later part of 2000). Among the systems discussed in earlier sections, four have currently received approval for clinical use, and only three systems are commercially available (Table 4). At the time this article was written (September 2004), there were nearly 483 facilities with more than 647 FFDM units in clinical practice as compared to 9,001 facilities with 13,625 SFM systems, with a certain overlap in the numbers of facilities (www.fda.gov/cdrh/mammography).


View this table:
[in this window]
[in a new window]

  		TABLE 4. Regulation and Accreditation of FFDM Systems

 
According to the Mammography Quality Standards Act of 1992 (MQSA) in the United States, a facility can provide clinical mammography services only after obtaining FDA approval. The FDA approval can be obtained through various accrediting bodies such as the American College of Radiology (ACR) (Reston, Va) (www.acr.org), which had been in the forefront of mammography quality control prior to enactment of the MQSA. It is one of the few accrediting organizations in the United States approved by the FDA to accredit facilities with FFDM systems. The current status of FFDM systems with regard to FDA approval and ACR accreditation is given in Table 4.

Clinical Trials

Currently, there is insufficient evidence to permit conclusions about the superiority of FFDM relative to SFM in screening for breast cancer. The benefits of digital mammography will be realized with the results of larger clinical trials. However, there have been a number of smaller clinical trials to date. One of the early trials, funded by the Office of Women’s Health of the U.S. Department of Health and Human Services, compared digital mammography with SFM in a diagnostic patient population with dense breasts. The study enrolled 201 women who underwent digital mammography at seven U.S. and Canadian medical centers (14,28) on three different FFDM systems (GE Healthcare, Lorad-Hologic, and Fischer Imaging). There was no significant difference in the cancer detection rate between the digital and SFM systems, but the interpretation of digital images depended on the lesion type (29).

A second federally funded study enrolled 4,945 women who underwent screening mammography on an FFDM system and an SFM system from the same manufacturer (GE Healthcare). No significant difference in cancer detection was observed between the FFDM and SFM systems. However, digital mammography resulted in fewer recalls than did SFM (30). The recall rate for FFDM (11.5%) was significantly lower than that for SFM (13.8%) (30). Results based on biopsy or 1-year follow-up indicate that the biopsy rate for FFDM (11.8%) was significantly lower than that for SFM (14.3%) (31); while FFDM had a lower sensitivity than SFM, the difference was not statistically significant.

A larger screening study designed to compare breast cancer detection sensitivity between SFM and FFDM during screening has completed its recruitment target of 49,500 women. The results from this study, called the Digital Mammographic Imaging Screening Trial (DMIST), funded by the National Cancer Institute and conducted by the American College of Radiology Imaging Network (ACRIN), are expected in 2004 or early 2005 (www.DMIST.org). The study is a paired design to compare FFDM systems from four manufacturers with SFM and was conducted at more than 30 sites in the United States and Canada. Reader variability may prove to cause great statistical variation, masking the differences between FFDM systems and between FFDM and SFM.

Phantom Studies

According to a recently published study by Berns et al (32), breast doses from FFDM are generally lower (by 10%–50%) than those of SFM. The radiation doses are especially lower for thicker breasts with FFDM compared to SFM systems. This is due to the inherently higher detection efficiency of digital detectors on most systems and the use of a harder x-ray beam at each breast thickness. Also, digital mammography provided better low-contrast lesion detection with phantoms than SFM (32,33). Another study with phantoms (34) found that there is the potential for radiation dose reduction by using different beam qualities with an FFDM system. The general consensus from these studies (32–35) is that the clinical use of digital mammography may generally improve image quality for equal or lower breast dose and with different beam qualities than SFM.

Challenges

There are a number of challenges with digital mammography technology. Since it is compared to SFM, the spatial resolution limitation is under increased scrutiny. The maximum spatial resolution achievable with a digital detector is 5–9 line pairs per millimeter, significantly lower than that obtained on SFM systems. However, the minimum pixel size needed for digital mammography is still under debate. Second, all three clinically approved FFDM systems in the United States (FDA approved) use a single receptor size. This poses a challenge for imaging breasts of all sizes and shapes, especially for imaging a large breast on a small digital detector, as this requires multiple exposures to cover the entire breast, then tiling partial breast images and reading the tiled image. Also, with current digital systems, images acquired on one manufacturer’s system cannot be easily displayed and processed on monitors of a different system. This makes it a challenge for users to read previous year images obtained on a different digital system. The requirements for archiving and storage of digital images can pose a challenge in terms of cost, especially to small freestanding breast screening centers.

Future Advances

Digital mammography is just the starting point for further advances in the diagnosis of breast cancer (36). The technology could enhance the applications, such as telemammography, image processing, and computer-aided diagnosis (CAD).

Telemammography
Telemammography can become very effective with digital mammography, since it allows underserved and geographically remote populations to access the latest in breast care. Currently, a number of feasibility studies are ongoing, including a field trial of mobile digital mammography to study Native American women in remote parts of Arizona. It is anticipated that telemammography, with concurrent digital acquisition and remote review of images, will allow complete evaluation of patients in a single visit at a remote location.

Image Processing
The digital format allows digital image processing to be applied to images. One such processing technique is contrast enhancement, a process whereby the contrast of different structures in the breast is altered to improve detectability. Also, image processing involves edge enhancement or smoothing the image and zooming in on a suspicious region in an image for better viewing. All these are possible without additional exposure to the patient.

Computer-aided Diagnosis
Several CAD systems are now clinically approved as second readers for SFM, in which x-ray films are digitized and processed for CAD reading, with certain potential for image degradation during digitization (37–39). Digital images facilitate CAD and remove the additional step of digitizing film.

Other advances in breast imaging currently under investigation include tomosynthesis (40–42), dual-energy imaging (43), digital subtraction mammography, and contrast-enhanced mammography (44,45).

Conclusions

Compared to SFM, digital mammography is still in its infancy. Despite the results from the early clinical trials indicating that there are no significant differences in the detection of breast cancer between the two, results such as significant reductions in recall rates and biopsy rate highlight the potential advantages of digital mammography in clinical care. Even though the initial costs of FFDM systems compared to SFM are high, the inherent advantages of digital mammography such as wide dynamic range, reduction in recall rates, potential for radiation dose reduction, increased patient throughput, ability to postprocess, and digital acquisition—which enables new applications to evaluate breast disease—will aid in the establishment of FFDM systems as a mainstay for breast evaluation. No matter what the results of larger clinical trials are, the technical developments that are occurring rapidly will ultimately demonstrate that the advantages of digital mammography will translate into improved patient care.

Footnotes

Abbreviations: a-Se = amorphous selenium, a-Si = amorphous silicon, CCD = charge-coupled device, FDA = Food and Drug Administration, FFDM = full-field digital mammography, SFM = screen-film mammography, TFT = thin-film transistor

References

  1. Jemal A, Tiwari RC, Murray T, et al. Cancer statistics, 2004. CA Cancer J Clin 2004; 54:8-29.[Abstract/Free Full Text]
  2. Nystrom L, Rutqvist LE, Wall S, et al. Breast cancer screening with mammography: overview of Swedish randomised trials. Lancet 1993; 341:973-978.[CrossRef][Medline]
  3. Hendrick RE, Smith RA, Rutledge JH, 3rd, Smart CR. Benefit of screening mammography in women aged 40–49: a new meta-analysis of randomized controlled trials. J Natl Cancer Inst Monogr 1997; (22):87-92.
  4. Laya MB, Larson EB, Taplin SH, White E. Effect of estrogen replacement therapy on the specificity and sensitivity of screening mammography. J Natl Cancer Inst 1996; 88:643-649.[Abstract/Free Full Text]
  5. Burrell HC, Sibbering DM, Wilson AR, et al. Screening interval breast cancers: mammographic features and prognosis factors. Radiology 1996; 199:811-817.[Abstract/Free Full Text]
  6. Sickles EA. Periodic mammographic follow-up of probably benign lesions: results in 3,184 consecutive cases. Radiology 1991; 179:463-468.[Abstract/Free Full Text]
  7. 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.[CrossRef][Medline]
  8. Yaffe MJ. AAPM tutorial. Physics of mammography: image recording process. RadioGraphics 1990; 10:341-363.
  9. Haus AG, Yaffe MJ. Screen-film and digital mammography: image quality and radiation dose considerations. Radiol Clin North Am 2000; 38:871-898.[CrossRef][Medline]
  10. Hendrick RE, Berns EA. Optimizing techniques in screen-film mammography. Radiol Clin North Am 2000; 38:701-718.[CrossRef][Medline]
  11. Bushberg JT, Siebert JA, Leidholdt EM, Boone JM. The essential physics of medical imaging Baltimore, Md: Lippincott, Williams & Wilkins, 2002.
  12. Feig SA, Yaffe MJ. Digital mammography. RadioGraphics 1998; 18:893-901.[Medline]
  13. Pisano ED. Current status of full-field digital mammography. Radiology 2000; 214:26-28.[Free Full Text]
  14. Pisano ED, Yaffe MJ, Hemminger BM, et al. Current status of full-field digital mammography. Acad Radiol 2000; 7:266-280.[CrossRef][Medline]
  15. Williams MB, Fajardo LL. Digital mammography: performance considerations and current detector designs. Acad Radiol 1996; 3:429-437.[CrossRef][Medline]
  16. White J. FDA approves system for digital mammography. J Natl Cancer Inst 2000; 92:442.[Free Full Text]
  17. Full-field digital mammography. TEC Bull (online) 2002; 19:1-4.
  18. Rowlands JA, Hunter DM, Araj N. X-ray imaging using amorphous selenium: a photoinduced discharge readout method for digital mammography. Med Phys 1991; 18:421-431.[CrossRef][Medline]
  19. Yaffe MJ. Direct digital mammography using a scanned-slot CCD imaging system. Med Prog Technol 1993; 19:13-21.[Medline]
  20. Tesic MM, Piccaro MF, Munier B. Full field digital mammography scanner. Eur J Radiol 1999; 31:2-17.[CrossRef][Medline]
  21. Muller S. Full-field digital mammography designed as a complete system. Eur J Radiol 1999; 31:25-34.[CrossRef][Medline]
  22. Vedantham S, Karellas A, Suryanarayanan S, et al. Full breast digital mammography with an amorphous silicon-based flat panel detector: physical characteristics of a clinical prototype. Med Phys 2000; 27:558-567.[CrossRef][Medline]
  23. Zhao W, Rowlands JA. X-ray imaging using amorphous selenium: feasibility of a flat panel self-scanned detector for digital radiology. Med Phys 1995; 22:1595-1604.[CrossRef][Medline]
  24. Zhao W, Ji WG, Debrie A, Rowlands JA. Imaging performance of amorphous selenium based flat-panel detectors for digital mammography: characterization of a small area prototype detector. Med Phys 2003; 30:254-263.[CrossRef][Medline]
  25. Kheddache S, Kvist H. Digital mammography using storage phosphor plate technique: optimizing image processing parameters for the visibility of lesions and anatomy. Eur J Radiol 1997; 24:237-244.[CrossRef][Medline]
  26. Jouan B. Digital mammography performed with computed radiography technology. Eur J Radiol 1999; 31:18-24.[CrossRef][Medline]
  27. Radiology fee schedule for film and digital screening and diagnostic mammography examinations—transmittal 1775, effective April 1, 2003. Medicare carriers manual part 3, publication 14–3. Centers for Medicare & Medicaid Services, U.S. Department of Health and Human Services 2002; 425-431.
  28. Cole EB, Pisano ED, Kistner EO, et al. Diagnostic accuracy of digital mammography in patients with dense breasts who underwent problem-solving mammography: effects of image processing and lesion type. Radiology 2003; 226:153-160.[Abstract/Free Full Text]
  29. Pisano ED, Cole EB, Major S, et al. Radiologists’ preferences for digital mammographic display: the International Digital Mammography Development Group. Radiology 2000; 216:820-830.[Abstract/Free Full Text]
  30. Lewin JM, Hendrick RE, D’Orsi CJ, et al. Comparison of full-field digital mammography with screen-film mammography for cancer detection: results of 4,945 paired examinations. Radiology 2001; 218:873-880.[Abstract/Free Full Text]
  31. Lewin JM, D’Orsi CJ, Hendrick RE, et al. Clinical comparison of full-field digital mammography and screen-film mammography for detection of breast cancer. AJR Am J Roentgenol 2002; 179:671- 677.[Abstract/Free Full Text]
  32. Berns EA, Hendrick RE, Cutter GR. Performance comparison of full-field digital mammography to screen-film mammography in clinical practice. Med Phys 2002; 29:830-834.[CrossRef][Medline]
  33. Berns EA, Hendrick RE, Cutter GR. Optimization of technique factors for a silicon diode array full-field digital mammography system and comparison to screen-film mammography with matched average glandular dose. Med Phys 2003; 30:334-340.[CrossRef][Medline]
  34. 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.[Abstract/Free Full Text]
  35. Gennaro G, Baldelli P, Taibi A, Di Maggio C, Gambaccini M. Patient dose in full-field digital mammography: an Italian survey. Eur Radiol 2004; 14:645-652.[CrossRef][Medline]
  36. Shah AJ, Wang J, Yamada T, Fajardo LL. Digital mammography: a review of technical development and clinical applications. Clin Breast Cancer 2003; 4:63-70.[Medline]
  37. Nawano S, Murakami K, Moriyama N, Kobatake H, Takeo H, Shimura K. Computer-aided diagnosis in full digital mammography. Invest Radiol 1999; 34:310-316.[CrossRef][Medline]
  38. Nishikawa RM. Computer-aided diagnosis complements full-field digital mammography. Diagn Imaging (San Franc) 1999; 21(9):47-51, 75.
  39. Roehrig J, Castellino RA. The promise of computer aided detection in digital mammography. Eur J Radiol 1999; 31:35-39.[CrossRef][Medline]
  40. Niklason LT, Christian BT, Niklason LE, et al. Digital tomosynthesis in breast imaging. Radiology 1997; 205:399-406.[Abstract/Free Full Text]
  41. Suryanarayanan S, Karellas A, Vedantham S, et al. Comparison of tomosynthesis methods used with digital mammography. Acad Radiol 2000; 7:1085-1097.[CrossRef][Medline]
  42. Dobbins JT, 3rd, Godfrey DJ. Digital x-ray tomosynthesis: current state of the art and clinical potential. Phys Med Biol 2003; 48:R65-R106.[CrossRef][Medline]
  43. Johns PC, Drost DJ, Yaffe MJ, Fenster A. Dual-energy mammography: initial experimental results. Med Phys 1985; 12:297-304.[CrossRef][Medline]
  44. Jong RA, Yaffe MJ, Skarpathiotakis M, et al. Contrast-enhanced digital mammography: initial clinical experience. Radiology 2003; 228:842-850.[Abstract/Free Full Text]
  45. Lewin JM, Isaacs PK, Vance V, Larke FJ. Dual-energy contrast-enhanced digital subtraction mammography: feasibility. Radiology 2003; 229:261-268.[Abstract/Free Full Text]



This article has been cited by other articles:


Home page
 RadioGraphicsHome page
M. Korner, C. H. Weber, S. Wirth, K.-J. Pfeifer, M. F. Reiser, and M. Treitl
Advances in Digital Radiography: Physical Principles and System Overview
RadioGraphics, May 1, 2007; 27(3): 675 - 686.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Mahesh, M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Mahesh, M.
Related Collections
Right arrow Breast (Imaging and Interventional)
Right arrow Physics and Basic Science