9.1 Technical Factors: kVp, mAs, AEC vs Manual, Target/Filter, Focal Spot, Grid, Mag, Labeling

Technique is the last controllable step before a mammogram becomes a medical record. On the ARRT Mammography examination (content specifications implemented September 1, 2025), Image Production treats kVp, mAs, automatic versus manual exposure, target/filter, focal spot, grid, magnification, and labeling as one working set. At the acquisition workstation (AWS), those choices decide whether the interpreting physician receives an image that can meet Enhancing Quality Using the Inspection Program (EQUIP) clinical-image criteria.

kVp: Subject Contrast Versus Penetration

Mammography uses a low peak kilovoltage (kVp) beam so photoelectric absorption can separate fat from fibroglandular tissue. Clinical settings cluster in the mid-20s to low-30s kVp. Generators can go wider; the exam tests the logic, not a vendor’s full scale.

• Lower kVp raises subject contrast and absorbed dose and lengthens exposure time.
• Higher kVp improves penetration of thick or dense breasts, shortens time (less motion), and lowers contrast.

On full-field digital mammography (FFDM) and digital breast tomosynthesis (DBT), display windowing can recover some grayscale, so a slightly harder beam is common. That does not make kVp irrelevant. Too hard, and dense tissue stays gray-on-gray after processing. Too soft, and automatic exposure control (AEC) runs long, the backup limit may trip, and motion unsharpness appears.

mAs: Quantity, Noise, and Time

Milliampere-seconds (mAs) set photon quantity. On film, mAs mainly changed optical density. On digital detectors, underexposure shows as quantum noise (grain) rather than a “light” image, and overexposure can still look acceptable after processing while delivering unnecessary dose.

After an AEC exposure, 21 CFR 900.12 requires the system to display the actual kVp and mAs used. Compare that mAs with what you expect for the compressed thickness: a 4 cm fatty breast that took a huge mAs, or a 7 cm dense breast that took almost none, is a technique problem even if the processed image looks pretty.

Long mAs (small focal spot, thick breast, low kVp, grid in) is the motion trap. If time is the problem, raise kVp or select a harder target/filter rather than stacking more mAs onto a moving patient.

Automatic Exposure Control

AEC (sometimes called an automatic exposure device) measures radiation reaching the detector and terminates the exposure. On most digital units it also selects kVp and target/filter so glandular tissue is adequately exposed. Two operator jobs remain:

1. Compressed thickness must be true. The paddle reports thickness into the AEC algorithm. A paddle that is not fully down, a thickness calibration that is off, or a flex paddle that reports the wrong height will pick the wrong spectrum.
2. The sensing region must sit under the tissue of interest, not under implant, hardware, open air, or a thin fatty tail. On systems with selectable AEC positions, the active cell is marked on the paddle; put dense tissue over it.

AEC is the default for contact screening and for most diagnostic views of intact breast tissue.

When Manual Exposure Is Required

AEC is inappropriate when the sensor cannot “see” a representative thickness of breast tissue. That is the exam distinction—not a preference for old-school charts.

Do not memorize one universal manual mAs. Start from a reasoned value for that thickness, then judge noise and contrast on the AWS.

Compression Thickness as a Technique Input

Thickness is not only a positioning goal. Digital AEC uses compressed breast thickness (and often a brief prepulse for composition) to choose kVp, filter, and mAs. If quality-control thickness accuracy is off, every automatic technique is biased. At the AWS, glance at the displayed thickness before you expose: a 3 cm-looking breast reported as 7 cm will get a beam that is too hard; the reverse looks noisy.

Adequate compression still matters for scatter, dose, overlap, and motion—those are image-quality criteria in the next section. Here the exam point is simpler: wrong thickness in → wrong technique out.

Target and Filter Selection

Classic screen-film spectra used molybdenum (Mo) or rhodium (Rh) targets with Mo or Rh filters. Characteristic peaks (Mo near 17–20 keV; Rh a few keV higher) and K-edge filtration shaped a narrow, high-contrast beam.

• Mo/Mo: highest contrast, best for thinner or fattier breasts; higher dose.
• Mo/Rh or Rh/Rh: more penetrating for thicker or denser tissue.
• Tungsten (W) with Rh, silver (Ag), or aluminum (Al) filters: common on FFDM/DBT; better penetration, lower dose, slightly less inherent subject contrast—which digital processing can partly restore.

Most digital units auto-select target/filter from thickness. You still need to know when to override: a very thin specimen or pancake breast may need the softest available combination for contrast; a thick, dense, or implant-in-place view needs a harder beam so the exposure finishes before motion. 21 CFR 900.12 requires the selected target to be indicated before exposure when more than one target is provided.

Focal Spot: Contact Versus Magnification

Mammography tubes provide a large (contact) focal spot—commonly on the order of 0.3 mm—and a small spot—commonly on the order of 0.1 mm—for geometric magnification. The system must indicate which spot is selected before exposure.

• Contact (grid-in) screening and most diagnostic views: large spot. Higher tube current, shorter time, less motion.
• Magnification: small spot. Geometric unsharpness grows with focal-spot size and object-to-image distance; the small spot keeps calcifications sharp on the mag platform.

Using the large spot on a mag stand is a classic sharpness fail. Using the small spot for routine contact views needlessly lengthens time.

Grids In Versus Out

A moving grid reduces scatter in contact geometry and is the default for craniocaudal (CC) and mediolateral oblique (MLO) views. MQSA requires magnification systems to operate with the grid removed. The mag air gap lets scatter miss the detector, so the grid’s contrast benefit is replaced and you avoid the extra time a grid would add to an already long small-spot exposure.

If grid lines appear, or a contact screening view was acquired with the grid out, that is a technique error—not a patient artifact.

Magnification Technique

Under 21 CFR 900.12, units used for noninterventional problem-solving must offer radiographic magnification, with at least one factor between 1.4 and 2.0. True geometric magnification is not the same as zooming a digital contact image.

Mag technique as a bundle:

1. Mag platform (increased object-to-image distance)
2. Small focal spot
3. Grid out
4. Spot compression when the indication is calcifications (thinner tissue, less scatter, shorter time)
5. Collimate to the area of interest for calcifications (less scatter); do not collimate so tightly on a mass work-up that landmarks disappear

Expect longer exposure times. Coach stillness. If the image is blurry, think motion first, then whether the large spot was left selected.

MQSA Image Identification (Labeling)

Each mammographic image must carry the following information in a permanent, legible, and unambiguous way, placed so as not to obscure anatomic structures (21 CFR 900.12(c)(5)):

1. Name of the patient and an additional patient identifier (medical record number is typical; date of birth is a weaker unique ID)
2. Date of the examination
3. View and laterality using accreditation-body standardized codes, placed near the axilla
4. Facility name and location — at minimum city, State, and ZIP code
5. Technologist identification
6. Cassette/screen identification (the screen-film cassette ID; on FFDM this is the detector/cassette identifier in the DICOM header)
7. Mammography unit identification if the facility has more than one unit

Those are the FDA fields. Do not treat kVp, mAs, filter, compression force, or thickness as MQSA-required labels. Those readouts are useful and often displayed; they are not the legal ID list. Wrong laterality, a missing second identifier, or view codes not near the axilla are labeling fails you correct before the study is released—not findings to “interpret around.”

AWS Habit Before You Send

Confirm AEC versus manual matched the clinical table, displayed thickness is plausible, focal spot and grid match contact versus mag, and every required ID field is present. That is technique evaluation, and it is what EQUIP later scores as examination identification plus exposure, contrast, and sharpness.