Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
Mammography Physics 101
A mammography system is a dedicated radiography system designed and optimized specifically for imaging the breast. Some systems still employ a cassette with a fluorescent screen and x-ray film, although most sites (>99.5%) in the United States use more modern devices that have a digital detector to record the x-ray image. As of March 1, 2020, there are 8696 (8671 with digital) certified mammography facilities with 21,563 accredited mammography systems. A total of 39,840,776 mammography procedures are performed annually ( https://www.fda.gov/radiation-emitting-products/mqsa-insights/mqsa-national-statistics ).
Image Acquisition
All mammography systems include an x-ray tube, filters, compression paddle, an antiscatter grid, and x-ray image receptor in a rotatable C-arm ( Fig. 21.1 ; Box 21.1 ). All systems have an automatic exposure control (AEC) system that can detect the lowest signal (densest region) behind the breast and provide feedback to adjust one or more technique factors (kV, filtration, tube current, imaging time; Box 21.2 ) to optimize image quality. The AEC may either use a separate x-ray sensitive device or use the image receptor itself to determine the signal level.
Box 21.1
Components of a Mammography System
| Generator | Converts 220 V (from the wall box) to the voltages required for system operation. In particular, the high voltage (25–45 kV) needed to energize the x-ray tube. |
| X-ray tube | A vacuum glass tube containing a rotating anode and small wire cathode. An accelerated electron beam from the cathode smashing into the anode generates x-rays that exit through the tube port. |
| Filter | A thin metal plate put into the x-ray field to filter out low energy x-rays. |
| Collimators | Lead or heavy metal strips near the x-ray tube port that move to limit the size of the x-ray field. |
| Compression paddle | A plastic semirigid plate that is used to immobilize and compress the breast. |
| Breast support table | Supports the breast and protects the sensitive detector underneath. |
| Antiscatter grid | A focused array of lead strips that preferentially block scattered x-rays while allowing unscattered primary radiation through to the detector. Some systems use two-dimensional honeycomb array rather than strips. |
| Imaging receptor | The x-ray sensitive detector. Usually a two-dimensional flat-panel array of detector elements. |
| Automatic exposure control (AEC) | The AEC is an electronic module, software control, or combination of both that provides feedback during the first few milliseconds of imaging to adjust the technique factors to achieve a target signal and/or noise level in the final image. The AEC may change the filter, kV, mA, and/or imaging time. |
| Acquisition workstation | The console from which the breast technologist controls image acquisition and determines image acceptability. |
| Picture and archiving system (PACS) | The network of computers, storage devices, and related software used to store and catalog images acquired in radiology departments. |
| Review workstation | A dedicated workstation with specialized high-resolution monitors for reviewing digital mammograms. |
Box 21.2
Technique Factors
| Tube potential, kV | The high voltage applied between the anode and cathode to accelerate the electron beam. |
| Tube current, mA | The measure of the current flow of electrons between the cathode and anode. |
| Tube current-time product, mAs | The total amount of radiation generated is proportional to the tube current (mA) multiplied by the imaging time (s). Radiation dose will be proportional to mAs. |
A certified breast technologist positions the breast between the compression paddle and the tabletop and applies compression. This reduces the effective thickness of the breast, reducing the radiation dose needed; spreads out the breast tissue so that some overlapping structures can be better seen; and reduces blurring due to patient motion. Note, however, that overcompression may lead to reduced imaging performance, although the mechanism for this loss in performance is not well understood.
The technologist hits the acquire button, and x-rays emitted from the x-ray tube port pass through the breast. Some are absorbed by the breast tissue while others pass through to the image receptor. The amount of x-ray absorption (or conversely transmission) depends on the x-ray energy, the compressed breast thickness, and tissue composition. The image captured by the image receptor is effectively an x-ray “shadow” of the tissue structure in the breast.
The x-ray tube ( Fig. 21.2 ) is powered by the generator, which applies a high potential of 25 to 45 kilovolts (kV) between the anode and cathode inside the x-ray tube. The tube potential (often referred to as the “kV”) is much lower than other radiography procedures (chest radiography or computed tomography [CT]). This is to take advantage of the differential x-ray absorption between soft tissue, which is increased at lower x-ray energies. This increases the contrast between the adipose (fatty) and fibroglandular (ductal tissue, stroma, and Cooper ligaments) tissues.
Early generators did not produce a constant tube potential and instead had the voltage rippling up and down. In those days, the tube potential was described by the maximum or “peak kV” (kVp). Because modern generators produce stable voltage levels, kV or kVp are used interchangeably.
The generator also controls the electron beam (the flow of electrons from cathode to anode), known as the tube current , which is measured in milliamperes (mA). Often, users simply state “mA” when referring to the tube current. Higher tube currents (more electrons) allow for shorter imaging times, but it causes a greater heating of the x-ray tube. The total tube output is the current-time product in milliampere-seconds (mAs).
The electron beam originates from the cathode, picks up kinetic energy from the kV applied between the anode and cathode, and the energetic electron beam slams into the anode. A tiny portion of that kinetic energy (~1%) is converted into energy that is released from the anode in the form of x-ray photons. The rest (~99%) is lost as heat to the anode. In a vacuum tube, heat can only dissipate slowly, and if an x-ray tube is used heavily, the tube may be damaged from overheating. Generally, the heating, or “tube loading,” of a mammography tube during normal clinical use is not a significant issue. However, when testing, images can be acquired very quickly and the amount of heat may build up. Overheating will damage the x-ray tube.
If the tube potential is 25 kV, then the x-ray photon emitted by the tube can have an energy up to 25 kiloelectron-volts (keV). If the tube potential is increased to 34 kV, then the x-ray photon will energies up to a maximum of 34 keV. An “electron-volt” (eV) is a convenient unit of energy equal to 1.6×10 -19 Joules (J). The x-rays emitted by an x-ray tube can have any energy from 0 up to the maximum specified by the tube potential. The distribution of energies is referred to as a “spectrum.” Example spectra are shown in Fig. 21.3 .
Target/Filter Combinations
X-rays emitted from an x-ray tube are polyenergetic, meaning that each x-ray photon from essentially zero keV to the maximum set by the tube potential. The energy distribution of the x-rays, or spectrum will be dictated by the composition of the “target” or anode. Generally, most anodes are composed of tungsten (W), although some anodes have special tracks made of molybdenum (Mo) or rhodium (Rh). A list the most common target/filter combinations for digital mammography is provided in Table 21.1 . The unfiltered x-ray spectrum is largely dominated by low-energy x-rays, which are generally highly absorbed by tissue (large organ dose) and provide little signal to the detector. As a result, a filter (a thin sheet of absorbing material, usually a metal of high purity and uniformity) is placed at the tube port to preferentially absorb most of the low-energy x-rays while letting the higher-energy x-rays through. The selection of the anode and filter are described as the anode/filter or target/filter combinations.
Table 21.1
Target/Filter Combinations Used in Digital Mammography
| Target (Anode) | Filter | kV Range (Typical) |
|---|---|---|
| Mo | Mo | 25–27 |
| Mo | Rh | 26–28 |
| Rh | Rh | 28–30 |
| Rh | Ag | 28–34 |
| W | Al | 28–32 |
| W | Ag | 29–36 |
Because of better dynamic range than screen-film mammography, digital mammography systems have moved toward higher-energy beams to provide greater signal at lower organ dose. To optimize these beams, common combinations from the screen-film era (Mo/Mo, Mo/Rh, Rh/Rh) have largely been supplanted by combinations such as Rh/Ag, W/Al, and W/Ag, though some systems still provide the former target/filter combinations for thinner breasts. For systems that offer contrast-enhanced mammography, an emerging technology, a copper filter may also be available for use.
The role of the filter is to remove the low energy x-rays. This is quantified by the “hardness” or penetrating capability of the x-ray beam. This is typically evaluated by a measure known as the half-value layer (HVL) thickness, which is the thickness of material (usually pure aluminum) needed to reduce the x-ray beam exposure by 50%. “Harder” beams will have thicker HVL values. As a rule of thumb, the MQSA regulations require that the HVL should exceed kVp/100 in mm of Al. For example, at 30 kV, the HVL must not be less than 0.30 mm (MQSA regulations Part B Table 2, therein, available at https://www.fda.gov/radiation-emitting-products/regulations-mqsa/mammography-quality-standards-act-regulations ). Harder beams provide more detector signal at reduced dose. However, the intrinsic contrast between adipose and fibroglandular tissue is also reduced. Key points on x-ray tubes and x-ray spectra are listed ( Box 21.3 ).
Box 21.3
X-Ray Tubes and X-Ray Spectra
- ■
The target or anode is the large rotating disk inside the x-ray tube on which the electron beam is focused.
- ■
Emission of x-rays from the target consists of bremsstrahlung (broad spectrum) and characteristic x-rays (narrow energy peaks, unique energy specific to anode material).
- ■
The filters are selected to remove low-energy x-rays and improve penetration of the x-ray beam at the breast.
- ■
The “hardness” of the x-ray spectrum is measured by the half-value layer (HVL).
- ■
Harder spectra can reduce dose and improve image signal-to-noise but reduce tissue contrast.
- ■
Harder spectra and higher kV are used for thicker and/or denser breasts.
Imaging Receptors
There are four main types of mammography systems based on their image receptor ( Box 21.4 ). The image receptor in screen-film mammography (SFM) is the screen-film cassette. The phosphor screen is the x-ray absorber, which emits visible light and exposes the film. The cassette is loaded into a film processor to extract the film and develop it in a chemical bath.
Box 21.4
Mammography Systems
| Screen-film mammography (SFM) | Older technology. Uses a phosphor screen as x-ray conversion layer, emitted light exposes film inside a removable cassette. Film must be developed in a chemical bath. Film mammograms are reviewed on a light box. |
| Computed radiography (CR) | Uses a cassette containing a storage phosphor. Phosphor is “read-out” using a laser scanner and digital image is recorded. Image quality is inferior to most digital mammography systems. |
| Digital mammography (DM) | Fully electronic image receptor. Usually integrated into the gantry. Usually, the image receptor is a flat-panel array that is 24 cm × 30 cm (9" × 12") in size. |
There are a variety of technologies employed for digital mammography. In one approach, computed radiography (CR), the film-screen cassette is replaced with a photostimulable phosphor plate inside a cassette. The phosphor plate absorbs the x-rays just like the screens used in SFM, but that energy is “trapped” by energizing the electrons in the phosphor crystals. The CR cassette is then taken to a reader or processor, inside of which a red laser light is scanned across the phosphor plate. The energy is released from the traps in the form of blue light, which is measured by a blue light sensitive photodetector.
The CR cassettes can be used in older SFM systems with little modification; however, studies have shown that the image quality is lower than other digital systems, higher doses are needed, and cancer detection rates can be impacted.
Most digital mammography (DM) systems ( Box 21.5 ) use a dedicated flat-panel detector that consists of a large two-dimensional array of detector elements that capture the signal from an x-ray (converter) layer. There are two types of x-ray detectors: indirect-conversion and direct-conversion systems ( Fig. 21.4 ). For indirect-conversion systems, the x-ray sensitive portion is a fluorescent material, or phosphor, that converts a portion of the x-ray energy to visible light photons. These photons are captured by the flat-panel detector array underneath it, which converts the light signal to an electronic charge that is subsequently read out and digitized into a numeric value. In the phosphor, the light photons are emitted in all directions, and careful selection and optimization of the phosphor material is required to limit the spread of the light before it reaches the detector. This light spread leads to blurring or unsharpness that can limit the detail visible in the final image. Direct-conversion detectors have a photoconductor plate directly deposited on the flat-panel array. Here, the x-ray energy is converted directly to electronic charges that are collected at the detector array. Unlike phosphors, however, the electronic charge is confined by an applied electric field and does not spread before it reaches the detector. As such, direct conversion detectors generally have resolution limits imposed only by the physical size of each detector element.
Box 21.5
Full-Field Digital Mammography Systems
In the literature, digital mammography is often referred to as full-field digital mammography (FFDM). This is to distinguish them from smaller format x-ray detectors that were first used in breast biopsy systems. FFDM or DM are both common abbreviations for digital mammography.
Every column in a flat-panel can be switched or “addressed” so that the signal can be read-out by a digitizer for the entire column at one time. The digitizer converts a voltage signal to a computer binary number. Most DM systems convert the signal to a value representing a binary number in 12 to 14 bits. This corresponds to a dynamic range of 0 to 4095 or 0 to 16,383 digital units.
Scatter and Grid
X-rays at mammographic energies interact with matter either by the “photoelectric effect” in which the photon energy is completely absorbed by the material or by “scattering,” in which the photon is deflected or scattered randomly in a direction that is different from the original primary beam. Scattered x-rays that reach the detector create a diffuse background signal that can degrade image quality. A similar effect is seen when driving in foggy conditions with high-beam headlights on, and the resulting glare makes it difficult to see objects ahead of the car.
Antiscatter grids ( Fig. 21.5 ) can be used to block the scattered x-rays preferentially. Usually, the grid is composed of strips of an attenuating material (usually lead) with spaces in between. The strips are angled toward the x-ray source such that the primary beam (unscattered) x-rays can pass through the spaces whereas any scattered x-ray photons traveling along different directions will be blocked by the strips. To avoid the grid appearing in the image, the grid is usually moved during imaging to blur out the strips uniformly. Special care must be taken to ensure that the grid motion is appropriate for the imaging time or a “grid artifact” will appear in the image. Grid artifacts generally require corrective action.
As described later, the antiscatter grid is not used when obtaining magnification views, in which scatter is reduced with the air-gap technique.
Heel Effect
The surface of the anode is aligned obliquely to the tube port ( Figs. 21.2 and 21.6 ), with x-rays emitted in straight lines from the anode and exit through the tube port and collimators. Some x-rays are emitted from the surface of the anode, but some are emitted deeper in the anode or must travel through different thicknesses of anode material. As a result, the x-ray spectrum and the number of x-rays are dependent on the exit angle of the x-ray. This leads to a nonuniform variation across the x-ray field ( Fig. 21.6 ), with higher field intensity toward the cathode than toward the anode. The effect of this nonuniform x-ray field on the image is mitigated through patient positioning. In mammography, the chest wall should be positioned toward the cathode side while the nipple should be positioned toward the anode.
Focal Spot and Magnification Views
In the x-ray tube, the electron beam is focused to a tiny area on the anode. This area, called the focal spot, is the location from which x-rays are generated. The focal spot has a finite size, for mammography tubes, these are nominally 0.3 mm for the large focal spot or 0.1 mm for the small focal spot that is used in magnification imaging. Because the focal spot is not a single point, a penumbra is created around objects in the image ( Fig. 21.7A ). Focal spot blurring (the penumbra in Fig. 21.7A ) can be minimized by using the small focal spot or by ensuring the object being imaged is close to the image receptor.
Unlike visible light, x-rays cannot be focused using lenses, so one cannot zoom in on an area of interest as would be done with a telephoto lens or a microscope. Instead, geometric magnification is used ( Fig. 21.7B ). Because x-rays travel in a straight line from the x-ray tube focal spot to the detector, the x-rays field is described as a diverging beam. To achieve geometric magnification, the object is moved closer to the x-ray tube and further from the image detector. The degree of magnification is equal to the source-image-distance (SID) divided by the source-object-distance (SOD). For example, if the object is moved halfway between the focal spot and image receptor, then the image would appear two times larger. If the object was placed one-third the distance from the focal spot to the receptor, then the image would appear three times larger. However, the penumbra effect from the finite-size focal spot becomes more apparent the further the object is from the detector. Because of this, the small focal spot is always used for magnification imaging.
In a magnification view, only a portion of the breast will be visible in the image. Smaller, spot compression paddles are used to immobilize and compress a portion of the breast. Spot compression views are used to characterize abnormalities, get better views of microcalcifications, or rule out tissue superposition. Note that less x-ray scatter reaches the image receptor in magnification mode due to the air gap, and the antiscatter grid is not used. MSQA regulations require that a mammography system must provide at least one fixed geometric magnification mode between 1.4× and 2×.
Radiation Dose
When x-rays interact with matter, they interact with the electrons in the atoms of the object. Either some (by scatter) or all (by photoelectric effect) of its kinetic energy is transferred to the electron. If that kinetic energy exceeds the binding energy of the electron, it can knock that electron from the atom resulting in an ion (a particle with a net charge) or break chemical bonds. The freed electron may have gained sufficient kinetic energy to interact with other electrons potentially liberating more electrons and breaking chemical bonds. At mammographic energies (~15–40 keV), this cascade of ionizations may cause several hundred or thousands of breaks.
If this ionization occurs near the DNA of the cell, the DNA molecule may be damaged either by these ionization events or by the reactive free ions (also called radicals) liberated by the x-ray energy. These free radicals can cause oxidative reactions that result in further DNA damage. This is in addition to the naturally present oxidative reactions and free radical generation during the normal metabolism of the cell. Fortunately, there are complex chemical repair mechanisms that continually repair and renew the cellular DNA. Only when these repairs fail is the DNA permanently altered (mutated). Even then, only certain mutations have the potential to transform the cell into cancer.
The amount of radiation delivered to an object or person is measured in grays (Gy), which is equal to 1 Joule deposited per kilogram of the (1 Gy = J/kg). Older references may indicate radiation dose in units of “rads” (100 rad = 1 Gy). Although radiation-induced cancer risk is well determined for high doses (from such studies as the Lifetime Survival Study after the atomic bomb attacks in Japan), the risk at low doses is difficult to determine. Conservatively, the risk at high dose is extrapolated to low doses using a linear no threshold (LNT) approach meaning that radiation risk is assumed to be linear with radiation dose and that there is no dose level below which the risk falls to zero. While the LNT model is commonly used to calculate the risk of radiation-induced cancer at high doses, its accuracy at low doses is unknown and some argue that it may overestimate risk.
For drugs, the dose is colloquially associated with the total amount of medicine administered (e.g., 400 mg of drug A). However, radiation dose is not an amount and is more like a concentration (of energy per unit body mass or J/kg). As such, if a woman receives 2 mGy from one mammogram on each breast, then she receives 2 mGy to each breast, not 4 mGy in total. However, if two repeated views were performed on one breast, then the total dose delivered to that breast would be 4 mGy. If the views are different (e.g., craniocaudal [CC] and mediolateral oblique [MLO] views) where the radiation is delivered to slightly different parts of the one breast, then the total dose to that breast would be roughly 4 mGy.
For mammography, we specifically refer to the dose delivered to the glandular tissue, which is the most susceptible to radiation (higher risk tissue) in the breast. Adipose tissue, stroma, and other tissues in the breast are generally considered to be more resistant (lower risk) to radiation exposure. As such, the dose to the breast is more properly referred to as the glandular dose. A common measure is the mean glandular dose (MGD) or average glandular dose (AGD), which are synonymous.
Radiation risk is inversely related to age; younger people are at higher risk of radiation-induced cancer, with the risk falling rapidly and tailing when approaching middle age. The radiation may induce cancer several years after exposure (5–30 years). For a typical MGD of 3.7 mGy for a two-view mammography examination, young women (age 20) may have a 4 in 100,000 risk of developing fatal radiation-induced cancer. At age 50, it drops to approximately 0.7 in 100,000 and down to 0.2 in 100,000 at age 70 ( ). In a similar analysis ( ), for 25 screens over a lifetime, the total combined risk of a fatal radiation-induced cancer was estimated to be 10 per 100,000 women.
In any risk analysis for an intervention, this should be compared against the benefit of the intervention. Lifetime risk of breast cancer is approximately 1 in 9 (or ~11,000 in 100,000) and the risk of a breast cancer fatality is approximately 1 in 30 (or ~3300 in 100,000). Estimates of the benefit of screening are between 15% and 45% based on various randomized controlled trials and observational screening studies. Assuming a moderate mortality reduction of 20% (660 deaths averted), then the benefit-to-risk ratio for 25 screens is approximately 66:1. Key points for radiation dose and risk ( Box 21.6 ).
Box 21.6
Radiation Dose and Risk
- ■
The mean glandular dose (MGD) is a common measure of radiation delivered to the breast.
- ■
Typical MGD are 1–6 mGy per mammogram, dependent on breast thickness and breast density.
- ■
The glandular tissue is the most sensitive tissue in the breast to radiation (x-rays).
- ■
Radiation risk is greatest in young women, such as those in their teens and 20 s.
- ■
Radiation risk is inversely proportional to age at exposure, and very low in screening aged women.
You're Reading a Preview
Become a Clinical Tree membership for Full access and enjoy Unlimited articles
If you are a member. Log in here