During the past three decades, there has been significant improvement in the equipment and image-recording systems for mammography. In addition to providing better images, these developments have resulted in a significant reduction in radiation dose. Mammography has evolved from more dedicated equipment and industrial film, through xeromammography, dedicated film-screen mammography, and now to full-field digital mammography (Fig. 2.1). With the importance of an emphasis on mammography in the role of early detection of breast cancer, it is of utmost importance that meticulous techniques be used. Factors that affect the image quality include equipment, image-recording system, processing, compression of the breast, and the technologist's skill in positioning the patient. Adherence to strict quality assurance guidelines is critical in mammography to maintain optimum image quality (1,2). The Mammography Quality Standards Act (MQSA) (3) was passed in 1992 and regulates mammography facilities in the United States. The entire mammography process is addressed through MQSA, but importantly, the quality of images is carefully evaluated.
The interpretation skills of the radiologist are limited by a suboptimal image. A poor-quality mammogram or poor positioning can account for many of the cancers missed by mammography (4,5), and technical errors should not be overlooked or accepted. For the radiologist to detect subtle signs of malignancy, high-quality images are a must. The emphasis of this book is on interpretation and pattern recognition, but it is important to review briefly the technical factors that affect the mammographic image.
Only dedicated units should be used for film-screen mammography. Dedicated units are those with special focal spots, target and filter materials, low kilovoltage, and compression devices designed to optimize the mammographic image at low radiation doses. Under no circumstances can nondedicated equipment produce images of similar quality to those performed with dedicated units.
There are three types of target materials used for mammography: molybdenum, rhodium, and tungsten. With molybdenum targets, 0.3-mm molybdenum filtration is used; the molybdenum targets are particularly well suited for mammography because of the low kiloelectron volt x-rays produced. The characteristic peaks for molybdenum are 17.9 and 19.5 keV, which provide high-contrast images for breasts of average thickness.
When the 0.3-mm molybdenum filter is used, the photons at energies greater than 20 keV are suppressed, and a larger number of low-energy photons are used in recording the images (6). The kilovolt peak (kVp) setting for molybdenum targets is generally at 25-28 kVp (6).
Alternative target/filter combinations for mammography include molybdenum/rhodium, rhodium/rhodium, and tungsten/rhodium. With tungsten targets, a beryllium window and minimal aluminum filtration are recommended. In comparison with a molybdenum target, even at low kVp settings, the tungsten target produces more high-energy photons, and the subject contrast is, therefore, lower (Fig. 2.2). Generally, settings of 22–26 kVp should be used for tungsten targets (A) (6,7,8). For most patients, a molybdenum target and a 0.06-mm molybdenum filter is automatically in place at low kVp settings; to penetrate thick dense breasts, the 0.05-mm rhodium filter is used at higher kVp settings (2).
Resolution is affected by the recording system's unsharpness, geometric unsharpness, and motion (9). The size of the focal spot of the tube is of particular importance in mammography because of the high resolution required for this work. To reduce geometric blurring, the focal spot
size and the distance between the breast and image receptor should be kept as small as possible, and the object-to-focal-spot distance should be maximized (10). The size of the focal spot becomes even more important in magnification mammography (6), where it is generally recommended that the measured focal-spot size be no greater than 0.3 mm (11) and preferably in the range of 0.1 to 0.2 mm (12,13).
Figure 2.1 HISTORY: Series of screening mammograms over 22 years.
MAMMOGRAPHY: Right CC view in 1985 (A), 1987 (B), 1996 (C), and 2006 (D) performed with the following equipment and techniques: A: Dedicated unit with a tungsten target, aluminum filter, and no grid. B: Dedicated unit with molybdenum target and filter and a grid. C: Dedicated unit with molybdenum target/filter, grid, and a newer film-screen combination with extended processing showing contrast improvement. D: Full-field digital unit with a tungsten/rhodium target/filter and a selenium detector showing excellent contrast and visibility of structures.
The use of grids in mammography improves image quality by reducing scatter, thus increasing contrast; there is a concomitant increase in radiation dose to the patient (14). Grids are of particular advantage in imaging the more dense, thick breast, in which more scattered radiation is present (15). However, because of the improvement in image quality with a grid, all routine mammography is performed with a grid. Mammography units have reciprocating grids with a ratio of approximately 5:1 (6). The grids used in mammography are thinner than conventional grids and contain carbon fiber interspace material for lower absorption. Typical reciprocating grids are composed of 16-µm lead strips separated by 300-µm carbon-fiber resin interspaces (7). The increase in the radiation dose with a grid is about two times that of a nongrid film, but this may be compensated for by using faster screen-film combinations. Grids are not utilized for magnification mammography because of the increased dose and length of exposure.
Dedicated mammography units should be equipped with a firm radiolucent compression device that forms a 90-degree angle with the chest wall. No rounded or curved edge compression devices should be used because the posterior aspect of the breast will not be adequately visualized (16). Proper compression of the breasts during mammography is extremely important in terms of producing an image of satisfactory diagnostic quality. Compression of the breasts is a very important factor in reducing scatter radiation, which degrades the image. With compression, there is spreading of the tissues apart, and small lesions are more easily identified within the parenchyma. The immobilization of the breasts decreases motion blurring, and the location of structures in the breast closer to the film receptor decreases geometric blurring. There is less variation in the density of the areas of breast with compression to a uniform thickness from nipple to chest wall. Importantly, the radiation dose to the breast is decreased with compression (7).
Image Recording System
It is important that a film-screen system specifically designed for analog mammography be utilized in order to
obtain the proper diagnostic quality of the images. With dedicated units and the common use of grids and magnification techniques, it is also important that the film-screen combination chosen have the lowest radiation dose while the quality of the images is maintained. Single-emulsion film is recommended for contact mammography because of the image degradation secondary to crossover in double-emulsion systems. In 1972, DuPont introduced the LoDose screen-film systems, which was followed by the DuPont LoDose 2 system. Kodak introduced the MinR system in 1976 and OM film in 1980. Since then, numerous film-screen combinations have been designed specifically for mammography. Some of these films have been specifically developed to be used in either standard 90-second or the extended 3-minute processors. The screens used for most routine mammography are single-back screens. The film and screen are in intimate contact, with the emulsion in contact with the top of the screen (17). Mammographic film-screen combinations have a much higher resolution than that of conventional radiography systems (2). Kimme-Smith et al. (17) found that the screens are of primary importance for good resolution, whereas contrast is affected more by type of film and processing (2,17).
Figure 2.2 Digital images of the American College of Radiology Accreditation phantom demonstrating the effect of target/filter combination on contrast. A was obtained using molybdenum/molybdenum target filter at 28 kVp and 97 mAs. B was obtained using tungsten/rhodium target/filter at 28 kVp and 83 mAs. The contrast is higher on the Mo/Mo image (A), and more objects are clearly visible (arrows).
Cassettes specifically designed for mammography are most commonly used, but polyethylene envelopes that are vacuum sealed have also been used. The cassette and its screen should be dusted daily to maintain proper quality control and to reduce dust artifacts. Weekly, wet cleaning of the screen is necessary to optimize image quality in reducing artifacts.
Quality control of film processing is of great importance in mammography. It is important to follow the manufacturer's recommendation for film processing in terms of the chemicals used, replenishment rate, development time and temperature. Using a temperature that is lower than that recommended by the manufacturer for mammography causes a loss in film speed and contrast (2), thereby necessitating a higher dose to produce films of satisfactory optical density (18). Many facilities have used extended-cycle processing for single-emulsion films (2,19,20). With extended-cycle processing, the film spends more time in the developer, thereby increasing the total processing time to 3 minutes. It is important that the processor be dedicated to mammography if extended-cycle processing is used. A reduction in radiation dose by 30% (2,19) and an increase (19) in contrast of 11% were found when Kodak OM1-SO177 film was tested in 3-minute rather than 90-second processing. Kimme-Smith et al. (21) found that extended cycle processing increased noise and decreased dose, but diagnostic capabilities were not compromised. More recently, newer films and screens have been developed that provide the higher contrast and lower dose afforded by extended cycle processing, yet the processing time for these films is 90 seconds.
Contrast is dependent on subject contrast (radiation quality and kVp), film contrast, film processing (darkroom, chemistry, development time and temperature), and scattered radiation (compression, grid) (22). There are a number of factors in the darkroom that affect radiographic contrast. The film should be processed according to the manufacturer's recommendations. The purity of the chemistry, the replenishment rate of the chemicals, and the temperature all affect the contrast. The time for processing also affects contrast significantly. Extended cycle processing has been used for mammography to increase the contrast of the images.
Another major effect on contrast is scattered radiation. The use of grids in mammography reduces scatter and increases contrast. The grid attenuates the primary radiation, though, and the dose is increased about twofold (23).
Figure 2.3 HISTORY: A 50-year-old woman with a lump deep in the left breast.
MAMMOGRAPHY: Left ML (A), CC (B), and MLO (C) views. No abnormalities are seen on the CC view, but on the ML view a questionable density (arrow) is seen positively. However, on the MLO view, there is a high-density nodular mass deep in the breast near the chest wall. This emphasizes the importance of performing the MLO view routinely in order to visualize the posterior aspects of the breast adequately.
HISTOPATHOLOGY: Infiltrating ductal carcinoma.
The parenchyma of the breast, not the skin, is the area that is of greater concern regarding exposure. Therefore, the absorbed glandular dose, not the skin entrance dose, is the more important measurement (6). For a 5-cm-thick average breast, the mean absorbed dose using OM film at 28 kVp with a molybdenum target is approximately 0.05 rad (6). Factors that affect radiation dose include breast-tissue composition and thickness, x-ray tube target materials, filtration, kVp, grid use, film-screen combination, and processing (2). With the development of dedicated mammography equipment and the improvement in
film-screen systems, there has been a tremendous decrease in the radiation dose from mammography over the last decades.
Figure 2.4 HISTORY: Screening mammogram with prior films for comparison on a difficult-to-position patient because of her size.
MAMMOGRAPHY: Current left MLO (A) view performed with a full-field digital unit shows the breast to be well positioned, especially for the large size. The pectoralis major muscle is visualized down to the level of the nipple. There is however, some overlap of the upper abdomen at the inframammary fold. A normal lymph node is noted in the axillary tail. Comparison with the prior study (B) shows that significantly more breast tissue has been included on the current study. The muscle, the node, and several centimeters of posterior breast tissue are visible on the well-positioned MLO versus the prior study.
IMPRESSION: Normal left mammogram with better positioning on the current MLO view.
Because of the resolution requirements for mammography, the technical developments to achieve full-field imaging have been challenging. Yet, full-field digital mammography has been developed and tested, and is clinically available. Other than the diagnostic improvements with digital compared with analog or film-screen mammography, as shown in the Digital Mammographic Imaging Screening Trial (DMIST) (24), there are distinct technical improvements as well.
With film-screen mammography, the various functions related to imaging are accomplished on a piece of film:
image acquisition, interpretation, and storage. With digital mammography, these three functions are separated, and each is optimized. The large dynamic range and improved contrast overcome some of the limitations of film-screen mammography. The dynamic range of digital mammography is approximately 1,600:1 versus 100:1 for film (25). The digital mammogram is particularly helpful because of increased contrast resolution in dense breast tissue.
Figure 2.5 Cross sections through the thorax show that the axillary tail of the breast may not be included on a ML projection (A), but on the MLO projection (B), the axillary tail and pectoralis major muscle are imaged.
The spatial resolution of digital mammography is less than that of film, with the current units having a resolution of 50- to 100-µm pixel size. The line pair resolution of film mammography translates to an approximate pixel size of 25 µm. However, because of the other favorable factors with digital imaging, fine detail observation is not compromised (26).
Digital detector development has included several different technologies. An early unit was composed of multiple small detectors (charge-coupled devices) stitched together to create a larger detector. Another technology is slot scanning, in which a long narrow CCD is scanned along with the x-ray beam over the breast. The GE unit was the first unit approved by the Food and Drug Administration and has a flat panel detector composed an amorphous silicone photodiode array with a spatial resolution of 100 µm. More recently, Siemens and Hologic have developed and received approval for units that are based on direct digital image capture on an amorphous selenium plate.
Simultaneous development of computerized radiography has occurred for mammography. Photostimulable phosphor plates are exposed to x-rays and develop a latent image. The plates are scanned by a laser beam, and the images are printed onto film. Like digital mammography, the phosphor plates offer a higher dynamic range than film mammography (26).
Digital images may be interpreted as hard copy or soft copy. With hard-copy interpretation, the images are laser printed onto film. Soft-copy interpretation involves interpretation at a high-resolution workstation. Because of the amount of information in one image, the images are not displayed at full resolution on a 2 K ÷ 2 K workstation. The images are enlarged to full resolution and panned to see the entire breast, or the magnification tool is used. A distinct advantage of interpretation at the workstation is that the image may be postprocessed. Window leveling, zoom, and magnification are all useful tools to improve observations and analysis.
Another important advantage of digital imaging is improved archiving of the images. Instead of a large film room and the challenges of maintaining it, digital images are archived in a PACS (picture archiving and communication system) or in a local and remote server. Because of the communication aspects of digital imaging, telemammography is possible and allows for remote interpretation.
Figure 2.6 Steps in positioning the patient for the MLO view. A: The technologist determines the angle of the obliquity of the pectoralis major muscle by lifting the breast medially and turning the image receptor to this angle. B: The Bucky is turned to the angle of obliquity of the pectoralis major muscle (35 to 60 degrees). The breast is placed over the Bucky with the arm draped behind the receptor. The technologist maintains the breast in a up-and-out position as she moves the compression device into position. C: The inframammary fold is open, and the breast is elevated as the compression is applied. D: Final position shows the entire breast and low axilla in the field of view.
Figure 2.7 Well-positioned MLO views show the pectoralis major muscles bulging forward and visualized as a triangle with the apex at the level of the nipple. The nipple is elevated, and the inframammary fold region is open; the breast is well compressed.
Figure 2.8 Positioning the patient for the CC view. (A) The technologist is using both hands to pull the breast forward over the receptor, and the receptor has been elevated to the elevated inframammary fold position. (B) On final positioning, the breast is pulled straight forward, and the medial aspect of the opposite breast is draped over the receptor.
Figure 2.9 Well-positioned CC views show the breasts positioned straight forward and not rotated. The pectoralis major muscles are seen bilaterally. The nipple is in profile, and the breast is well compressed.
Figure 2.10 ML. Positioning the patient for the ML view. The breast is compressed from the medial aspect, and the degree of obliquity is 90 degrees. The axillary tail is not as included in the field of view as it is on the MLO view.
Most authors agree that two views should be performed for routine mammography (27,28). The craniocaudal (CC) view and the mediolateral oblique (MLO) view (29) are recommended as standard projections. Before the determination of the MLO projection, the views for mammography were a CC and a mediolateral (ML). With the advent of MLO positioning, more posterior tissue could be included in the field of view (Fig. 2.3). Additional views may be necessary to evaluate specific areas within the breast (30,31,32), and the techniques for positioning the patient for these various views are described in this chapter. For all views, it is of utmost importance that the breast be compressed properly.
Mediolateral Oblique View
When the patient is positioned properly, the MLO view will demonstrate the pectoralis major muscle and the entire breast, including the inferior portion and the axillary tail, on one film (Fig. 2.4). The concept of the MLO view is that the breast is compressed at the same angle of obliquity as the pectoralis major muscle transverses the chest wall. In compressing the muscle in this plane, the technologist is more able to pull the muscle forward over the cassette; therefore, the posterior breast tissue is also pulled forward (Fig. 2.5). The pectoralis major muscle has a triangular shape, with the apex at the level of the nipple. In general, the parenchyma should not extend to the posterior edge of the image on the MLO view. Instead, the parenchyma should be separated from the chest wall edge of the image by the retroglandular fat (33).
The MLO view is performed by angling the receptor and compression device, 45 to 60 degrees in a caudal direction from the vertical position. The level of angulation is determined by the orientation of the pectoralis major muscle on the chest wall. The technologist determines this angle by visualizing the edge of the patient's
pectoralis major muscle in the anterior axillary line, and she moves the equipment to this angle. The degree of obliquity depends on the patient's body habitus; a thin patient needs a steeper oblique; a heavy, short patient may be examined with a lesser degree of obliquity. The patient stands with the ipsilateral arm elevated to no more than 90 degrees over the receptor. The receptor is placed behind the breast, high into the axilla, and the patient may gently hold the cassette holder or handle.
Figure 2.11 HISTORY: A 39-year-old woman with a palpable mass left breast.
MAMMOGRAPHY: Left MLO (A) and CC (B) views show a lobular, high-density mass with indistinct margins. On the ML (C) view, the lesion appears to be located higher than on the MLO, confirming a medial location.
IMPRESSION: Highly suspicious for carcinoma.
HISTOPATHOLOGY: Medullary carcinoma.
NOTE: Lesions that are in the medial aspect of the breast will appear higher on the ML than on the MLO view.
The positioning of the breast is improved when the technologist uses the natural breast mobility in positioning. The lateral and inferior aspects of the breast are more naturally mobile than are the medial and superior aspects (34). Therefore, in positioning the patient for the MLO view, the lateral aspect of the breast and the pectoralis major muscle are displaced medially and anteriorly and then positioned over the image receptor. When properly positioned, the MLO view includes nearly all of the breast tissue (34) (Figs. 2.6 and 2.7).
In this position, the pectoralis major muscle can be most easily pulled forward from the chest wall, enabling greater visualization of the posterior aspects of the breast. The patient should not be allowed to tense the arm, because this will also tighten the pectoral muscle and prevent the breast from being pulled forward easily. All aspects of the lateral side of the beast and the axilla should be in contact with the cassette; if the opposite breast is in the radiographic field, the patient should press it up and against the chest wall. If this is not done, the opposite nipple can project into the radiographic field, simulating a nodule (35). The breast is pulled forward
and up as the compression device is applied. In doing so, the inframammary fold is open, and the posterior and inferior aspect of the breast is visualized. Vigorous compression must be applied without allowing the breast to sag on the cassette.
Figure 2.12 HISTORY: A 56-year-old woman with a history of multiple benign breast biopsies, presenting now with nystagmus and neurologic findings suggesting a paraneoplastic syndrome. An outside mammogram had been interpreted as nonsuspicious for a primary breast cancer.
MAMMOGRAPHY: Enlarged right MLO (A) and magnification (2X) CC (B) views and right XCCL (C) and ML (D) views from a needle localization. The breast is moderately dense. There are two clusters (arrows) of microcalcifications, irregular in contour and separated by approximately 2 cm in the right lower outer quadrant (A and B). These calcifications were considered highly suspicious for malignancy, and needle localizations were performed before excisional biopsy. The XCCL (C) and ML (D) views from the needle localization (C and D) show a spiculated 8-mm mass (arrow) deep in the lower outer quadrant near the chest wall.
HISTOPATHOLOGY: Infiltrating ductal with multicentric intraductal carcinoma.
NOTE: The ML view may demonstrate a lesion near the chest wall in the inferior aspect of the breast better than the routine MLO view. In this case, the demonstration of the mass was serendipitous, because the final films for the localization included a standard ML view, on which the lesion was seen.
The CC view is a standard transverse view of the breast. The CC complements the MLO view by visualizing the far medial and posterior aspects of the breast. The CC view includes most of the parenchyma except for the far lateral and posterior regions of the breast. On the CC view, the central and subareolar areas are well compressed and well depicted.
In considering the natural mobility of the breast, the inferior aspect is elevated during the positioning of the CC view. With this maneuver, the superior and posterior tissue are included in the field of view and are better visualized. The patient stands facing the mammographic unit with the head turned away from the breast being examined. The patient should lean slightly forward, and the technologist places one hand under the breast, elevating it. The image receptor is placed at the level of the
elevated intramammary fold. The breast is pulled forward over the cassette with both hands. Skin folds and wrinkles should be smoothed out; rotating the shoulder slightly back helps to smooth out the skin folds at the anterior axillary line. The nipple should be in profile if at all possible. The medial aspect of the opposite breast is draped over the edge of the receptor (Fig. 2.8). The films are marked in the axillary region, and firm compression is applied to the breast.
Figure 2.13 Positioning of the patient for the left LM view. This is a 90-degree lateral view in which the compression plate is placed on the lateral side of the breast. The x-ray beam is directed from the lateral to medial direction. The medial aspect of the breast is closest to the receptor, thereby optimizing imaging of medial structures.
On a well-positioned CC view, the distance from the edge of the film at the chest wall to the nipple should be no more than 1 cm greater or less than the distance measured from the nipple to the anterior border of the pectoralis major muscle on the MLO (the posterior nipple line [PNL] measurement) (34). In a study of 1,586 CC mammograms, Bassett et al. (36) found that in 79% of cases, the measurement of the PNL on the CC view was within 1 cm of the PNL on the MLO view. The authors (36) also found that the pectoralis major muscle was seen on 32% of the CC views (Fig. 2.9).
Occasionally, a lesion is seen only on the CC view and not on the MLO view. A rolled CC view can be obtained by sliding the superior aspect of the breast medially or laterally and the inferior aspect of the breast in the opposite direction (31). If the “lesion” persists, the direction of its movement relative to its position on the standard CC view indicates its relative vertical position in the breast.
In addition to the routine CC and MLO views, supplementary views may be used in the evaluation of a mammographic or palpable abnormality. A tailored examination with views selected to evaluate the potential abnormality is a diagnostic mammogram. The purpose of the additional views is multifactorial: to determine if a potential lesion is real, to evaluate its position in the breast, to evaluate its characteristics, and sometimes to search for other occult lesions. These additional views include the reverse obliques (lateral-medial oblique, superoinferior oblique), 90-degree lateral views (ML, lateral medial), exaggerated CC views (lateral, medial), rolled CC views, cleavage view, off-angle obliques, spot compression, magnification, axillary tail and axillary views, tangential views, and implant displaced views.
True Lateral (90-Degree) Views
The ML view may not demonstrate the posterior and axillary portion of the breast in entirety; therefore, the MLO view is recommended instead for the routine examination (29). The ML view, however, is essential in localizing a lesion and may also be of help in differentiating a true lesion from superimposition of glandular tissue. With the cassette placed against the lateral aspect of the breast and compression applied from the medial direction, the ML view is a true sagittal view (Fig. 2.10). The ML view may also be extremely useful in demonstrating a lesion located high in the upper inner quadrant, an area sometimes not included on the MLO view, or a lesion deep near the chest wall in the inferomedial or inferolateral aspect of the breast. The ML view can be used to locate a lesion demonstrated on an MLO view but not seen on a CC view. A lesion that is lower in position on the ML view than on the MLO view is located laterally; a lesion that is higher in position on the ML view than on the MLO view is located medially (Figs. 2.11 and 2.12).
Figure 2.14 HISTORY: Screening mammogram on a patient with a history of multiple cysts and fibrocystic changes.
MAMMOGRAPHY: Right MLO (A) and bilateral CC (B) views show the breasts to be heterogeneously dense. There is an indistinct mass seen at the posteromedial aspect of the right breast on the CC view only (arrow). Because of the position of the lesion, a lateromedial LM view (C) was obtained to try to locate the lesion on the sagittal plane and to place it closer to the image receptor. The mass is visible on the LM view superiorly (arrow) and is also visible on the spot CC (D) view.
IMPRESSION: Dense indistinct mass located at 2 o'clock posteriorly on the right breast, demonstrated on the LM view.
HISTOPATHOLOGY: Invasive ductal carcinoma.
Figure 2.15 Positioning for the LMO view shows the breast compressed from the lateroinferior to the mediosuperior direction. The angulation is usually the same degree of obliquity as is used for the MLO view.
The lateral medial (LM) view is also a 90-degree lateral view; however, in this case the breast is compressed from the lateral aspect. The x-ray beam is directed from the lateral to the medial direction. This view places the medial aspect of the breast closest to the receptor, which is particularly useful for improving the visibility of lesions that are located medially (Figs. 2.13 and 2.14).
The advantage of performing the lateromedial oblique (LMO) view is to image lesions located far medioposteriorly that are seen on the CC view only or to image palpable lesions in the inner quadrant that are not seen on mammography. To position the patient for this view, the tube and cassette holder are tilted 45 to 60 degrees toward the contralateral breast (Fig. 2.15). The medial aspect of the breast to be examined is placed against the cassette holder, and the ipsilateral arm rests over the receptor. The breast is elevated so that it does not droop, and compression is applied from the lateral direction. This view is also used for patients who are very kyphotic or in patients with a pacemaker or a port located in the upper inner quadrant. The view may also be helpful to demonstrate lesions located medially and not seen on the MLO (Fig. 2.16).
The superoinferior oblique (SIO) view, like the LMO, is used for kyphotic patients or those with pacemakers and central lines. The view is an oblique projection, but instead of the x-ray beam being directed medially to laterally, it is in a superolateral to inferomedial direction. The compression plate is placed against the lateral and
superior area, and the breast is compressed from this direction (Fig. 2.17).
Figure 2.16 HISTORY: A 44-year-old gravida 3, para 2, abortus 1 woman with no palpable findings, for screening.
MAMMOGRAPHY: Right MLO (A), CC (B), XCCM (C), and LMO (D) views. Although on the routine MLO view (A) no abnormality is seen, the edge of a well-circumscribed mass (arrow) is present far medioposteriorly on the CC view (B). An XCCM view (C) demonstrates the lesion more clearly. To identify the exact position of the mass, the LMO view (D) was performed and showed the mass to be in the upper inner quadrant.
NOTE: The LMO is particularly useful for demonstrating lesions located medially, near the chest wall.
Exaggerated Craniocaudal Views
The tissue in the extreme lateroposterior or medioposterior aspects of the breast may not be visualized in entirety on the routine CC view. When a lesion is found on the routine MLO view deep in the breast and is not seen on the CC view, an exaggerated lateral or medial CC (XCCL or XCCM) view will be of help in defining the location of the abnormalities. For the XCCL view, the lateral aspect of the breast is placed forward on the cassette in the CC position (Fig. 2.18). For the axillary tail of the breast to be in good contact with the film, the patient must lean backward slightly, keeping the ipsilateral arm extended over the top of the cassette. For the XCCM view (Fig. 2.19), the patient is rotated anteriorly, extending her chest forward, with the far medioposterior aspect of the breast being imaged. If the lesion is located high in the upper inner quadrant, it may be necessary to elevate the cassette holder and compress the uppermost aspect of the breast.
The exaggerated views are used to determine the location in two projections of a lesion seen only on the MLO posteriorly. The XCCL is performed first because more parenchyma and more lesions, especially cancers, are located in the upper outer quadrant than elsewhere (Figs. 2.20,2.21,2.22,2.23).
If a lesion is located at the chest wall on the MLO and not seen on the XCCL, it is presumably located medially. In this case, the XCCM view is performed to extend the
field of view posteriorly at the medial side of the breast (Fig. 2.24).
Figure 2.17 Positioning the patient for the SIO view. The degree of obliquity is the same as for the MLO view, but the breast is compressed from the lateral side. The direction of the beam is from the superolateral to inferomedial direction. This is used for kyphotic patients and those with pacemakers.
In addition to the XCCM, the cleavage view is performed for possible medial and posterior lesions (Fig. 2.25). For the cleavage view, both breasts are placed over the image receptor with the cleavage in the center of the field. Usually a manual technique, rather than phototiming is used, because the photocell is not covered completely by breast tissue.
For lesions that are seen only on the MLO and not demonstrated on the ML or CC, step obliques or off-angle obliques can be most helpful. The concept is that the angle of obliquity is changed about 10 degrees more than and less than the obliquity the MLO. If the lesion is visible on these additional obliques, it is real. The displacement of the lesion on the additional views also is used to locate it. Lesions that are located in the lateral aspect of the breast appear lower on the higher angle of obliquity (approaching the 90-degree lateral) (Fig. 2.26).
Most mammographic units have a smaller compression paddle that can be used for spot compression. The area of concern, identified on the standard view, is spot compressed, with the surrounding normal tissue pushed away by the compression device (Fig. 2.27). The technologist must estimate the location of the lesion in the breast from the initial images, and care must be taken to be certain that the area of concern is included in the spot field. Spot compression is particularly useful for the evaluation of the borders of nodules and for focal densities that may represent either true lesions or overlapping tissue (Figs. 2.28 and 2.29). Another use of spot compression is in the evaluation of a palpable nodule not clearly seen on mammography. The technologist rotates the palpable lesion into tangent with the beam and spot compresses the area (30). Magnification
may be combined with spot compression, particularly in the evaluation of a small nodule or calcifications.
Figure 2.18 On the standard CC view (A and B), a lesion located posterolaterally may not be included on the image, but by rotating the patient for the XCCL view (C), the lesion projects into the radiographic field. For this position, the breast is rotated medially so that the far lateral and posterior aspect of the breast is included in the field of view.
Figure 2.19 On the standard CC view (A and B), a lesion located posteromedially may not be included on the image, but by rotating the patient forward for the XCCM view (C), the lesion projects into the radiographic field. The breast is rotated laterally so that the far medial and posterior aspect is included in the field of view. The opposite breast is draped on the image receptor.
Figure 2.20 HISTORY: A 65-year-old gravida 4, para 3, abortus 1 woman for screening.
MAMMOGRAPHY: Right MLO (A), CC (B), and XCCL (C) views. The breast is moderately dense. There is a focal spiculated lesion in the upper aspect of the breast on the MLO view (A), but this lesion is faintly seen (arrow) on the CC view (B). The XCCL view (C) is performed to demonstrate the location and the appearance of the lesion (arrow). The XCCL view is performed first if a lesion is seen only on the MLO, because more carcinomas occur laterally than medially. If a lesion is not found on the XCCL position, then the XCCM view is performed.
IMPRESSION: Spiculated lesion located posteriorly in the upper outer quadrant, highly suspicious for malignancy.
HISTOPATHOLOGY: Infiltrating lobular carcinoma.
Figure 2.21 HISTORY: Screening mammogram on a 70-year-old woman.
MAMMOGRAPHY: Right MLO (A) and CC (B) views show heterogeneously dense tissue. There is a small mass at the superior margin of the parenchyma (arrow) seen on the MLO view only. Because the lesion was not seen on the CC view, an XCCL view (C) was obtained. This projection shows that the lesion is located far laterally and posteriorly (arrow).
IMPRESSION: Suspicious mass in the right upper-outer quadrant, demonstrated on the XCCL view.
HISTOPATHOLOGY: Invasive ductal carcinoma.
When a possible lesion is seen on the CC view but not the MLO view, rolled CC views are useful to determine if the lesion is real and to locate it in the sagittal plane. The rolled CC views are performed by positioning the patient for the CC view and before compressing the breast, slightly displacing the tissue. For the rolled CC medial view, the upper pole of the breast is displaced medially before compression. For the rolled CC lateral view, the upper pole of the breast is displaced laterally (Fig. 2.30).
The interpretation of the rolled views requires an understanding of the displacement for each view. If the “lesion” is still apparent on the rolled CC views, it is a real finding. Its position in the breast is determined by the direction of displacement. If the lesion appears more medial on the rolled medial and more lateral on the rolled lateral, it is in the upper half of the breast (Fig. 2.31). If it moves in the opposite direction from the roll, it is in the inferior pole.
Although the axillary tail of the breast and the inferior aspect of the axilla are seen on the MLO view, it may be necessary to obtain an additional view to evaluate the upper axilla. The tube and cassette are angled 45 to 60 degrees from the superomedial to the inferolateral direction. The patient is turned 15 degrees away from the mammographic unit, and the ipsilateral arm is placed at a 90-degree angle to the breast. The cassette is placed
behind the ribs with the edge just above the patient's humeral head, and the patient leans slightly backward to optimize contact with the cassette (Fig. 2.32). The axilla is the only area to be included on the film. Vigorous compression is not obtained because of the many structures interposed. Imaging of this region is typically performed to evaluate palpable masses or to evaluate a mass or masses seen in the superior aspect of the MLO image (Figs. 2.33 and 2.34).
Figure 2.22 HISTORY: A 62-year-old woman with a palpable mass in the right breast.
MAMMOGRAPHY: Right MLO view (A) shows a large dense round mass at the chest wall. The lesion was not visible on the CC view (B). The XCCL view is performed to rotate the posterolateral aspect of the breast forward and to try to image the lesion in a transverse projection. On the XCCL view (C), the spiculated lesion is noted laterally and posteriorly.
IMPRESSION: Highly suspicious for carcinoma.
HISTOPATHOLOGY: Invasive ductal carcinoma.
The tangential view is an ancillary view that is used primarily to (a) assess a palpable lump, (b) visualize the area of the tumor bed after lumpectomy and radiotherapy, and (c) confirm that calcifications are dermal. For palpable lumps or for lumpectomy scars, the area of interest is placed in tangent to the x-ray beam and is compressed in this orientation (Fig. 2.35).
For possible dermal calcifications, a localization procedure is performed. The localization compression plate is placed over the surface of the breast in which the calcifications are thought to be located, and an image is obtained. The coordinates of the calcifications are identified, and a BB is placed over these coordinates. The localization plate is then removed, and the skin surface beneath the BB is placed in tangent. If the calcifications are dermal, they project at the skin (Fig. 2.36).
Imaging the Patient with Implants
The presence of breast implants for augmentation poses particular difficulty in obtaining adequate mammographic
images of the surrounding parenchyma. Standard MLO and CC views using manual techniques are necessary in order to image the posterior aspects of the breast and the implant.
Figure 2.23 HISTORY: A 43-year-old woman for screening mammography.
MAMMOGRAPHY: Left CC view (A) shows heterogeneously dense tissue and grouped microcalcifications laterally (arrow). On the left XCCL magnification views (B), groups of pleomorphic microcalcifications (arrow) are present. There is also a small spiculated mass located posteriorly (arrowhead) that was not visualized on the standard CC view.
IMPRESSION: Highly suspicious for multifocal carcinoma.
HISTOPATHOLOGY: Invasive ductal, nuclear grade 2, and ductal carcinoma in situ.
Figure 2.24 HISTORY: A 48-year-old woman with a palpable 4.5-cm mass in the right upper-inner quadrant.
MAMMOGRAPHY: Right MLO (A), ML (B), and XCCM (C) views. The MLO view is not of optimum quality because the posterior aspect of the breast and the pectoralis muscle are not visualized completely. There is a relatively well-defined mass in the midportion of the breast. Posterior to this mass, there is an ill-defined area of increased density extending to the edge of the film. It is important that if one identifies an area of increased density such as this, the breast posterior to it should be evaluated with additional views. On repositioning the breast (B), the 3.5-cm high-density mass is identified. XCCM spot (C) demonstrates the spiculated lesion in the far posteromedial aspect of the breast.
IMPRESSION: Carcinoma of the breast.
HISTOPATHOLOGY: Infiltrating ductal carcinoma.
Figure 2.25 Positioning for the cleavage view shows both breasts over the receptor. Phototiming is not used because the automatic exposure control is not positioned over parenchyma.
Figure 2.26 HISTORY: Screening mammogram on a 62-year-old patient with a history of multiple cysts and benign biopsies.
MAMMOGRAPHY: Left MLO view (A) shows scattered fibroglandular densities and multiple round masses consistent with cysts. In the far posterior aspect of the left axillary tail is a small indistinct density (arrow) that was not clearly evident on prior studies. On the ML view (B), the density was not evident. Off-angle obliques obtained at angulation greater than and less than that of the MLO (C, D)again show the small density, which persists and has the same shape. It appears to be located more inferiorly on the oblique at 76 degrees (C) versus the oblique at 52 degrees, indicating that it is located laterally. This is the same concept as the movement in the position of the lesion on the MLO versus the 90-degree lateral or ML. Lesions located laterally appear lower on the 90-degree lateral than they do on the MLO view. The lesion is confirmed on the XCCL (E) view as being located laterally (arrow).
IMPRESSION: Small mass in the axillary tail confirmed with off-angle MLO positioning.
HISTOPATHOLOGY: Invasive ductal carcinoma.
Figure 2.27 Positioning for the spot compression in the MLO projection shows the small compression paddle over the area of interest.
Figure 2.28 HISTORY: A 56-year-old gravida 10, para 10 woman for screening. Bilateral MLO (A) view. There is asymmetry between the breasts, with irregular increased density (arrow) being noted in the right subareolar area.
MAMMOGRAPHY: Right CC (B), and spot compression (C) views. The area (arrow) appears somewhat spiculated on the CC view (B). However, a spot compression of the area (C) demonstrates clearly the spiculated mass beneath the nipple. The appearance is highly suspicious for carcinoma.
HISTOPATHOLOGY: Infiltrating ductal carcinoma.
Figure 2.29 HISTORY: A 52-year-old woman for screening mammography.
MAMMOGRAPHY: Right MLO (A) and CC (B) views show scattered fibroglandular densities. There is an irregular dense mass located in the upper outer quadrant. On spot compression magnification ML (C) and XCCL (D) views, the mass is better seen to be an irregular lesion with indistinct margins, and it is associated with pleomorphic microcalcifications.
IMPRESSION: Highly suspicious for malignancy.
HISTOPATHOLOGY: Invasive ductal carcinoma with mucinous features.
Figure 2.30 Rolled CC views. For a rolled CC medial view (A), the breast is placed in a CC position, and the superior pole is displaced medially as the inferior pole is displaced laterally. For a rolled CC lateral view (B), the breast positioned in a CC orientation, and the superior pole is rolled laterally as the inferior pole is rolled medially. Lesions that are located in the upper aspect of the breast move medially on the rolled medial and laterally on the rolled lateral views. Lesions located inferiorly move in the opposite direction from the roll.
Eklund et al. (37) described a modified positioning technique that allows for better compression and imaging of the anterior parenchymal structures. In positioning the patient for the modified technique, the technologist first palpates the anterior margin of the implant. The breast tissue anterior to the implant is pulled forward and placed over the cassette. For the CC view, the cassette tray is raised more than for standard positioning, and the inferior edge of the implant is displaced behind the cassette tray. As the compression plate is brought down, the technologist gently pulls the parenchyma forward and displaces the implant posteriorly, guiding the superior margin of the implant behind the descending compression device. For the MLO view, the same procedure is used; the lateral edge is placed behind the cassette holder, and the compression is guided over the medial aspect of the implant (Figs. 2.37 and 2.38).
Well-positioned implant studies show the implant to be included on the standard views, with some visualization of the axillae and axillary tails. On the modified positioning or implant displacement views, the implants are not visible, and the anterior breast tissue is well compressed and visualized (Fig. 2.39).
Direct radiographic magnification results in improved sharpness and detail (38) compared with conventional mammography. There is an improvement in the effective
resolution of the recording system, reduction of the effective noise, and reduction of scattered radiation (10). The breast is elevated from the Bucky by placing it on a magnification stand (Fig. 2.40). Small focal spots are used for magnification; because of this, there is a decrease in tube output and increase in exposure time (10). This produces an increase in blur from patient motion and an increase in radiation dose (10). The kVp is increased to compensate for the longer exposure needed for proper optical density. Magnification is of help in defining the borders of mass lesions, morphology, and number of microcalcifications (Figs. 2.41,2.42,2.43), and in determining the existence of multicentric tumor (39). There is an increase in radiation dose to the breast of 1.5 to 4 times that of conventional mammography (38).
Figure 2.31 HISTORY: A 56-year-old woman recalled for a left breast density.
MAMMOGRAPHY: Left CC view (A) shows a small indistinct mass located medially (arrow). This mass was not seen on the MLO view. Rolled CC views medially (B) and laterally (C) were obtained and show that the lesion persists (arrows) and displaces with the superior pole of the breast. On spot compression (D), the indistinct margins and high density of this small mass are noted.
IMPRESSION: Small mass in the upper inner quadrant, highly suspicious for carcinoma.
HISTOPATHOLOGY: Invasive ductal carcinoma.
Figure 2.32 Positioning of the patient for the axillary tail view shows the upper aspect of the breast and the axilla included in the field of view.
Figure 2.33 Proper positioning and contrast on the axillary view demonstrate an enlarged axillary node.
Figure 2.34 HISTORY: An 82-year-old woman for screening mammography.
MAMMOGRAPHY: Bilateral MLO views (A) show normal parenchyma and some prominent lymph nodes in the axillae. Axillary views (B) include more tissue in the axillae and better demonstrate the adenopathy. The inferior breast is not compressed or positioned in such a way to be evaluated on these views. The axillae are emphasized and are included in the field of view.
IMPRESSION: Axillary views showing mild adenopathy bilaterally.
Figure 2.35 Positioning for the tangential view is tailored to the area of interest. A BB is placed over the area of interest, and that area is placed in tangent to the x-ray beam (A, B).
Assessment of Image Quality
The role of the radiologist in mammography begins with an assessment of image quality. This must be performed routinely before image interpretation and is critical to the detection of breast cancers. Image quality is assessed, in addition, during the accreditation process for mammography facilities. Clinical image evaluation during the accreditation process includes assessment of positioning, compression, image quality, artifacts, and labeling (22). Imaging of the American College of Radiology Accreditation phantom is used to test the system performance as well by evaluating both contrast and resolution.
Figure 2.36 HISTORY: A 42-year-old patient referred for evaluation of calcifications seen on screening mammography.
MAMMOGRAPHY: Left MLO (A) and CC (B) views show a small cluster of calcifications (arrows) located at 12 o'clock, somewhat superficially. On the CC magnification view (C), the calcifications are clustered and very well demarcated, suggesting that they are dermal. On the tangential ML view (D), the calcifications are clearly depicted within the skin (arrow) seen best on the enlarged image (E).
IMPRESSION: Tangential view demonstrating calcifications to be dermal.
Figure 2.37 Positioning the patient with implants. (A) The standard MLO view with the implant in the field of view. (B) The technologist is displacing the anterior edge of the implant posteriorly as she is pulling the parenchyma forward. (C) The implant-displaced MLO view with only the parenchyma included in the field of view. Much of the parenchyma is included in the image with only the posterior regions not being visible.
Figure 2.38 Positioning the patient with implants: the CC view. (A) The breast including the implant is included in the field of view. (B) The technologist is displacing the implant posteriorly as she is pulling the parenchyma anteriorly and placing it over the receptor. (C) The implant-displaced CC view includes only the anterior parenchyma in the field of view.
Figure 2.39 Normal implants and mammography. (A) Standard MLO views showing the saline implants in the field of view. (B) Standard CC views showing the saline implants included. (C) Implant-displaced MLO views show the subpectoral implants displaced posteriorly. The parenchyma anterior to the implants is well compressed and visualized. (D) Implant-displaced CC views show the pectoralis muscles and the breast parenchyma included and only an edge of the right implant visible. A scar marker indicates a prior benign biopsy site.
Figure 2.40 Positioning the patient for spot magnification. The breast is elevated away from the receptor on the magnification stand(arrow), increasing the object-to-receptor distance and creating an air gap. Spot compression is applied over the area of interest.
Clinical image evaluation should be performed by the technologist and the radiologist on each mammogram. Evaluation of image quality includes an assessment of technical factors, such as exposure, contrast, noise, and sharpness. An underexposed image may have occurred because of incorrect position of the photocell, and such underexposure greatly compromises the detection of masses or calcifications in dense parenchyma. Optimizing contrast in mammography is critical to the visualization of subtle differences in tissue densities. Noise or radiographic mottle is affected by the number of x-rays used to produce the image. Fewer x-rays are associated with increased quantum mottle and decreased ability to visualize fine calcifications (22). Sharpness is related to the geometry of the focal spot, motion unsharpness, and screen unsharpness from poor film screen contrast. Proper compression of the breast is important in reducing motion unsharpness.
The images should be free from artifacts. The types of artifacts visible on mammography may include those related to the patient, such as overlying structures (chin, ear, hair, eyeglasses) (Figs. 2.44 and 2.45), or to substances on the skin (deodorant or powder). Artifacts on film-screen mammography images are different from those on digital mammography. On film-screen mammography, common artifacts that relate to film and cassette handling include dust, fingerprints, and focal fog.
Lastly, the assessment of image quality includes the evaluation of positioning. Factors such as proper positioning and compression on the MLO view to include the posterior tissue and proper positioning on the CC view to include the medial tissue are part of this evaluation. In patients in whom the posterior tissue is very thick relative to the nipple area, uniform compression
may not be attainable. Obtaining additional CC and MLO views of the anterior aspects of the breasts in both projections can optimize compression and imaging of both regions of the breast. For lesions that are clinically evident, a tailored examination is necessary to be sure that the palpable area is included in the field of view (Fig. 2.46).
Figure 2.41 HISTORY: A 72-year-old gravida 6, para 4, abortus 2 woman for screening.
MAMMOGRAPHY: Right ML (A), enlarged (1.5X) CC (B), spot magnification (2X) (C), and specimen (D) views. There is scattered glandularity present. Extensive vascular calcifications are seen (arrowheads, A and B). On the initial CC view (B), there are microcalcifications (arrow) that appear to project beyond the lumen of the calcified vessel. These calcifications (arrow) are better evaluated with the spot compression magnification view (C), on which they are clearly displaced away from the calcified vessel(arrowhead). Their contour is slightly irregular, and they are, therefore, of moderate suspicion for malignancy. These were biopsied after needle localization (D), and the specimen film demonstrates their clustered nature and slightly irregular morphology.
IMPRESSION: Moderately suspicious calcifications demonstrated on spot compression magnification view.
HISTOPATHOLOGY: Intraductal carcinoma.
Figure 2.42 HISTORY: A 54-year-old patient recalled for an abnormal screening mammogram.
MAMMOGRAPHY: Right magnification ML (A) and CC (B) views show highly pleomorphic clustered microcalcifications (arrow) at 12 o'clock superimposed on dense parenchyma. The magnification views demonstrate very clearly the shapes and borders of the microcalcifications.
IMPRESSION: Highly suspicious for malignancy.
HISTOPATHOLOGY: High-grade ductal carcinoma in situ.
The viewing conditions are also evaluated through MQSA and are critical to the optimization of visualization of breast abnormalities. For film-screen mammography, high-intensity view boxes with shutters to reduce extraneous light are required. The ambient light in the reading room should be kept to a minimum. Likewise, with digital mammography, high-luminance workstations and low ambient light are used for image interpretation. Intensity windowing at the workstation has been shown to improve detection of simulated calcifications in phantoms (40). Kimme-Smith et al. (41) found that the detection of microcalcifications in dense breasts was significantly affected by viewing conditions.
Even with good techniques, 5% to 10% of breast cancers are not detected by mammography. It is of utmost importance that the radiologist maintain high standards of quality, be certain that excellent positioning is performed, and correlate the clinical examination with mammogram in determining that the region of interest is included on the film. By maintaining these standards, the number of cancers not detected by mammography will be kept to a minimum.
Figure 2.43 HISTORY: Patient recalled from screening for evaluation of microcalcifications.
MAMMOGRAPHY: Left MLO (A) and CC (B) views show a cluster of microcalcifications at 10 o'clock. A magnification CC (C) view demonstrates the morphology and number much more clearly. Magnification mammography is very important in the assessment of microcalcifications and in defining their possible etiologies and management.
Figure 2.44 Artifact during positioning the patient for the CC view from her eyeglasses overlying the field of view.
Figure 2.45 Coned-down right CC view showing a striated density (arrow) at the chest wall. This is the patient's hair extending forward between the tube and receptor and overlying the field of view.
Figure 2.46 HISTORY: A 55-year-old gravida 2, para 2 woman with a large tender mass in the right axillary tail.
MAMMOGRAPHY: Right MLO (A) and repositioned right MLO (B) views. On the initial view (A), the breast is moderately dense, and no suspicious abnormalities are seen. The patient did have a palpable lump in the axillary tail; for this reason, the technologist repositioned her to bring the mass forward into the field of view (B). More of the axillary area is seen on this view (B), but the breast itself is not compressed as well on the standard MLO (A). A large, partially circumscribed mass with microlobulated borders is present in the right axillary tail and is most consistent with breast carcinoma, although a metastatic node is another consideration.
IMPRESSION: Carcinoma, right axillary tail, demonstrated on a tailored MLO view to demonstrate the palpable mass.
HISTOPATHOLOGY: Poorly differentiated adenocarcinoma.
NOTE: It is critical that the technologist palpate any masses noted by the clinician or patient and be certain that the area of palpable concern is included in the field of view.