8.1 Mammographer QC: Phantom Image, Compression Thickness, Visual Checklist
Key Takeaways
- The 2018 ACR Digital Mammography QC Manual is an FDA alternative standard; facilities may instead follow manufacturer QC, and those frequencies can differ.
- ACR 2018 phantom image quality is weekly using an ACR-approved digital mammography phantom compressed to about 5 daN (about 12 pounds).
- Digital phantom pass criteria are the two largest fibers, three largest speck groups, and two largest masses, with no clinically significant artifacts.
- Compression thickness indicator and visual checklist are monthly ACR 2018 mammographer tests; indicated thickness is commonly required within ±5 mm of actual thickness.
- ARRT items on mammographer QC target purpose, frequency, equipment and procedure, performance criteria, evaluation and documentation, and corrective action.
Why Image Production D starts with the mammographer's tests
The ARRT Mammography exam treats quality control (QC) as a working skill, not a policy slogan. Under Image Production D, you need a detailed understanding of every mammographer QC test listed in the content specifications and a basic understanding (purpose and frequency) of medical physicist tests. Both lists are referenced to the 2018 ACR Digital Mammography Quality Control Manual. Exam questions on technologist tests typically ask purpose, frequency, equipment and procedure, performance criteria, how you evaluate and document the result, and what you do if the test fails.
This section covers the three tests that live on the mammography unit itself in the ACR 2018 program: phantom image quality, compression thickness indicator, and visual checklist. Monitor QC, repeat analysis, viewing conditions, compression force, and detector calibration are in 8.2. Physicist tests are in 8.3 and 8.4.
Manufacturer QC versus the 2018 ACR alternative standard
The Mammography Quality Standards Act (MQSA) at 21 CFR 900.12 requires a QC program that is substantially the same as the image-receptor manufacturer's recommendations, or an FDA-approved alternative. The 2018 ACR Digital Mammography QC Manual (with the digital breast tomosynthesis supplement) is that alternative for 2D and DBT. FDA still expects manufacturer QC for contrast-enhanced mammography (CEM) applications even when the rest of the unit is on the ACR program.
If the question stem says the facility follows manufacturer QC, do not force ACR weekly/monthly answers onto it. Manufacturer manuals vary by vendor, model, and software version. If the stem says ACR 2018, use the frequencies in this chapter. A unit should not mix the two programs. ACR training is explicit that a medical physicist must complete a transition survey before technologists drop manufacturer tests and start ACR 2018 tests.
MQSA also expects a designated QC technologist. Other qualified people may acquire a phantom image, but the QC tech owns method, review, and the decision to call service or the physicist.
Weekly phantom image quality
Purpose. The phantom is a stand-in for a 4.2 cm compressed breast of average density. Scoring it asks whether the imaging chain — automatic exposure control (AEC), beam quality, detector, processing, and the display you score on — still shows objects that represent fibers (spiculations/lines), speck groups (calcifications), and masses, and whether artifacts would interfere with diagnosis.
Frequency (ACR 2018). Weekly. This is the only routine weekly unit test in the ACR digital program (computed-radiography cassette erasure is extra if a facility still uses CR). Screen-film daily processor QC is not the digital default. Manufacturer programs may still require phantom images more often; follow the manual in force.
Equipment and procedure. Use an ACR-approved digital mammography phantom purchased from a listed vendor. It is larger than the old small accreditation phantom and covers most of the detector, which is why it is better at catching artifacts at the periphery. ACR accreditation support (phantom testing, revised 2026) describes the technique:
- Place the phantom so its chest-wall edge is flush with the chest-wall edge of the image receptor. A gap here is the same geometric miss as a posterior nipple line that is short on a patient.
- Compress to approximately 5 daN (about 12 pounds). That is phantom technique, not the 25–45 pound clinical power-drive range tested semiannually (8.2).
- If the unit has a selectable AEC sensor, center it under the wax insert.
- Use the clinical technique for a 4.2 cm average breast (AEC on, typical target/filter and kVp).
- On DBT-capable units, acquire the modes the QC program requires (typically 2D and tomosynthesis).
- Score from the acquisition workstation when that display is adequate. Record kVp, mAs, filter, and compressed thickness on the form.
Performance criteria. For the ACR digital mammography phantom, pass requires no clinically significant artifacts and visualization of the two largest fibers, the three largest speck groups, and the two largest masses. The old small phantom used four fibers, three speck groups, and three masses. ACR designed the two keys to be equivalent in difficulty; mixing them is the highest-yield trap on this topic. Artifacts that are clinically significant fail the test even if every object is visible.
Evaluation and documentation. Score with the ACR rules: a fiber counts when the full length is seen; a speck group counts when enough specks in the group are seen; masses are scored as complete or not. Deduct when an artifact mimics or hides an object. Date, initial, and keep the form. FDA retention guidance keeps weekly QC documentation for the previous 12 weeks; monthly records are kept at least until the next inspection cycle is closed. Follow the facility policy and MQSA.
Corrective action. A failed phantom is an image-quality failure of a weekly test. MQSA 21 CFR 900.12(e)(8) requires identifying the problem and taking corrective action before further examinations that would use the failed component for daily and weekly tests, for dosimetry, and for other-modality (digital) QC. Repeat the phantom after the fix. If it still fails, stop and involve service or the medical physicist. Do not invent a homemade 30-day grace period for a failed phantom.
Worked example: objects pass, artifact fails
You see the two largest fibers, three speck groups, and two masses, but a broad banding artifact crosses the detector. The object score is not a pass by itself. Clinically significant artifacts fail the phantom test. Do not image patients on that unit until the artifact is explained, corrected, and the phantom is repeated.
Monthly compression thickness indicator
Purpose. The console thickness is not a courtesy number. It feeds AEC logic, displayed organ dose, and comparison with prior studies. A 1 cm error can change automatic technique and make serial comparisons dishonest.
Frequency (ACR 2018). Monthly.
Equipment and procedure. Compress an object of known thickness (the phantom itself or a measured spacer) using a specified force. ACR training uses about 5 daN, the same order as phantom imaging. Read the indicated thickness and compare it with the known value. Some facilities use the older small phantom as a reproducible spacer; tape-measure methods are less reproducible.
Performance criteria. Training and review materials commonly cite ±5 mm of actual thickness. If the facility's current ACR or manufacturer form prints a different limit, that form is the action limit you follow.
Worked example. A 40 mm spacer reads 46 mm. The error is 6 mm, outside ±5 mm. Document the fail, remove the unit from clinical use until the indicator is repaired or recalibrated, and repeat the test. Do not keep imaging while you wait for a convenient service window if the QC form treats this as a fail that affects examinations.
Monthly visual checklist
Purpose. Phantom scoring will not tell you that a paddle is cracked, a cable is frayed, a lock does not hold, the receptor cover is dirty, or the DBT assembly does not travel through its range. The visual checklist is the mechanical and cleanliness survey of the room.
Frequency (ACR 2018). Monthly.
Procedure. Use the checklist in the QC manual: cleanliness of the breast support and paddles; smoothness of C-arm and compression motion; compactness of compression (paddle stays parallel and fully contacts the phantom); indicator lights and displays; face shield; cable condition; and DBT movement if the unit has tomosynthesis. Mark each item pass or fail. Do not initial a line you did not look at.
Corrective action. A failed item that affects safety or image formation takes that function out of service until it is repaired. A cracked paddle is not a next-week problem. Document what you found, who was notified, and the retest after repair.
How the three unit tests fit together
| Test | ACR 2018 frequency | Purpose | Typical performance idea |
|---|---|---|---|
| Phantom image quality | Weekly | Objects + artifacts on a 4.2 cm stand-in breast | 2 fibers, 3 speck groups, 2 masses; no significant artifacts |
| Compression thickness indicator | Monthly | Displayed thickness matches true thickness | Often ±5 mm |
| Visual checklist | Monthly | Mechanical, cleanliness, and safety items phantom scoring misses | Required items pass |
If the facility is still on manufacturer QC, rewrite the frequency column from that manual. The purposes stay the same; the calendar may not.
Exam traps for 8.1
- Weekly phantom, not daily. Daily darkroom and processor tests were screen-film.
- Digital phantom pass is 2 / 3 / 2, not 4 / 3 / 3.
- Phantom compression is about 5 daN, not 25–45 pounds.
- Flush chest-wall placement matters as much as object counts.
- Always read whether the stem is ACR 2018 or manufacturer QC.
A facility uses the 2018 ACR Digital Mammography QC Manual for 2D and DBT. How often must the QC technologist acquire and score the ACR digital mammography phantom on each unit?
Using the ACR digital mammography phantom (not the small screen-film-era phantom), which object-visibility combination meets typical pass criteria if there are also no clinically significant artifacts?
Under the 2018 ACR Digital Mammography QC Manual, how often is the compression thickness indicator checked, and what is the commonly cited performance window versus actual thickness?