PET/CT ACR Accreditation Phantom QC

The ACR PET phantom is the accreditation test that proves a PET/CT scanner recovers accurate standardized uptake values (SUV) and resolves small structures—turning "the pictures look fine" into a measured, defensible claim about quantitative accuracy. ^{1, 4} It fills a cylindrical phantom to a known activity concentration, scans it with the laboratory's own clinical protocol, and checks that background SUV sits near 1.0, that hot cylinders recover contrast, and that cold rods are visible.

Introduction

PET/CT is a quantitative imaging test. Unlike a chest radiograph, where interpretation is qualitative, a PET report often turns on a number: the SUV of a lesion, its change since the last scan, or whether it crosses a response threshold. ^{2, 3} That number is only as trustworthy as the scanner's calibration and image quality. A PET/CT whose calibration has drifted will report SUVs that are systematically high or low, and a clinician comparing a follow-up scan to a baseline can misjudge whether a tumor is responding. ^{10, 11}

The ACR PET phantom is how a laboratory demonstrates—not assumes—that its quantitative accuracy is intact. ⁴ By filling a phantom to a concentration that should read as an SUV of 1.0, scanning it exactly as patients are scanned, and measuring the result, the medical physicist converts a vague sense that "the scanner is working" into documented evidence that SUV, contrast recovery, and uniformity all meet defined tolerances.

This guide explains what the ACR PET phantom measures, the physics of SUV and contrast recovery behind the test, how it relates to the NEMA and AAPM acceptance-testing frameworks, and how it fits into a defensible PET/CT quality program. DRPS delivers this testing as part of its PET/CT and nuclear medicine physics and accreditation support services across Florida, Maryland, Virginia, Washington DC, California, and Nevada.

Topic Explanation

What the ACR PET phantom is

The ACR PET phantom is a cylindrical accreditation phantom containing hot cylinders of decreasing diameter, cold rods arranged in sectors, and a large uniform background compartment. ^{4, 5} The physicist fills the background and the hot inserts to prescribed activity concentrations, lets the short-lived positron emitter mix, and scans the phantom with the same acquisition and reconstruction protocol used for clinical whole-body PET. The resulting images are then analyzed three ways: ⁴

• Background SUV — a large region of interest (ROI) in the uniform compartment, which a correctly calibrated scanner should read near 1.0.
• Hot-cylinder contrast and recovery — SUV measured in the hot cylinders (including 25 mm and 16 mm inserts) and their ratios, which test how faithfully the scanner reproduces activity in progressively smaller structures. ⁴
• Cold-rod visibility — whether the cold-rod sectors can be resolved, a low-contrast and spatial-resolution check. ⁵

The test is deliberately end-to-end. It does not isolate a single component; it exercises the entire chain—dose calibrator, clock synchronization, detector calibration, normalization, attenuation and scatter correction, reconstruction—and asks whether the number that comes out the far end is right. That is exactly the chain that determines a patient's SUV. For the daily calibration steps that feed into this, see our guide to PET/CT daily QC and calibration, and for the meaning of the SUV itself, PET SUV quantification.

Why background SUV should equal 1.0

SUV is defined as the measured activity concentration in tissue normalized to the injected activity distributed over the body: ²

where $C_{\text{img}}$ is the image-derived activity concentration (for example, in kBq/mL), $A_{\text{inj}}$ is the injected activity, and $m$ is the body mass (with 1 mL of tissue taken as 1 g). ² The ACR phantom is filled so that the uniform background concentration equals the total phantom activity divided by the phantom volume—precisely the denominator of the SUV equation. By construction, a perfectly calibrated scanner should therefore measure a background SUV of 1.0. ⁴ A measured background that drifts away from 1.0 is a direct, quantitative signature of a calibration error somewhere in the chain, most often the dose calibrator, the scanner calibration factor, or a clock mismatch between them.

Key Technical Principles

Two quantitative ideas underpin the ACR PET phantom: SUV accuracy (does the scanner get the concentration right?) and contrast recovery (does it get the contrast right in small objects?). The table contrasts the ACR accreditation phantom with the NEMA NU 2 image-quality phantom used at acceptance, because they answer different questions. ^{4, 7}

Contrast recovery: a worked example

Small hot structures never reach their true contrast in a PET image because the scanner's finite spatial resolution blurs activity across the object boundary—the partial-volume effect. ² The contrast recovery coefficient (CRC) quantifies how much of the true contrast survives:

where $C_{\text{hot}}$ and $C_{\text{bkg}}$ are the measured hot-structure and background concentrations and $a_{\text{true}}$ is the true activity-concentration ratio at fill time. ^{7, 10} Suppose the NEMA IQ phantom spheres are filled to a true hot-to-background ratio of 4:1. If the 37 mm sphere measures a ratio of 3.7:1:

A small 10 mm sphere in the same scan might recover only 40–60% of its contrast because partial-volume blurring is proportionally larger. Tracking CRC across sphere sizes is how acceptance testing characterizes resolution-limited quantitation, and it is the same physics that limits how faithfully the ACR hot cylinders recover SUV. ^{7, 10} For the underlying resolution and its causes, see PET spatial resolution and positron range and the PET partial-volume effect.

The acceptance-testing tolerances behind the QC

The ACR annual phantom test lives inside a larger acceptance-testing framework. AAPM Task Group 126 defines a manufacturer-independent set of PET/CT acceptance tests, and a recent evaluation of a clinical scanner using TG-126 reported the tolerances the phantom program is built to protect: the accuracy of count-rate corrections, expressed as a background SUV, within about 10% of the expected value of 1.0; image uniformity within about ±5%; and PET-to-CT image registration within about 1 mm. ⁸ The same study measured sensitivity of roughly 19–20 counts per second per kBq and a peak noise-equivalent count rate near 2,174 kcps—system descriptors that the NEMA NU 2 standard defines and that anchor a scanner's baseline. ^{7, 8} These acceptance numbers set the expectation the routine ACR phantom then confirms over the scanner's life.

Clinical Impact

Phantom QC is invisible to patients, but its failure is not. When a PET/CT's calibration drifts, every SUV in every report shifts with it—and because SUV is used to judge treatment response, a calibration error can convert a responder into a non-responder or the reverse. ^{10, 11} A multi-scanner study of FDG PET/CT showed that unharmonized SUV differences between scanners are large enough to change how a patient's response is classified. ¹¹ The ACR phantom, by pinning background SUV to a known value, is the guardrail that keeps those errors from reaching the report.

The clinical demands on SUV accuracy are concrete. Response criteria such as PERCIST compare a lesion's SUV—normalized to a liver reference region—against defined percentage thresholds to declare progression or response. ³ Those thresholds only mean something if the SUV scale is stable and accurate; a scanner whose background SUV reads 1.15 instead of 1.0 biases every measurement by 15% and can push a borderline case across a decision line. Harmonization programs such as EANM/EARL exist precisely because quantitative comparability across sites and scanners cannot be assumed—it has to be measured and maintained. ^{1, 9}

Contrast recovery matters just as much for small lesions. A sub-centimeter node may recover only half its true SUV, so a physician who does not account for partial-volume effects can underestimate disease. Understanding the CRC-versus-size behavior measured on phantoms tells the reader how much to trust the SUV of a small structure—and tells the physicist how reconstruction choices trade resolution against noise. ^{7, 10}

There is also a practical, high-yield reason the ACR phantom is the workhorse of PET quality: it catches the single most common cause of SUV error, a broken cross-calibration between the dose calibrator and the scanner. SUV depends on two independent activity measurements—the dose calibrator that assays the patient's injection and the scanner's own calibration factor—and if those two devices disagree, every SUV is biased even though the daily detector QC looks perfect. A phantom filled to a known concentration is the only routine test that exercises both measurements together and exposes the mismatch. That is why a background SUV drifting from 1.0 so often traces back not to the PET detectors but to the dose calibrator, the clock, or the calibration-factor entry—components a picture-quality check would never flag. ^{4, 8}

Practical Optimization Tips

A defensible ACR PET phantom test follows a disciplined workflow.

1. Control the fill and the clock

The single largest source of phantom error is the activity measurement and timing. Measure the fill activity accurately, record the assay time precisely, and confirm the scanner's clock agrees with the dose calibrator's clock. Improved filling procedures that substitute accurate time and volume measurements for a difficult direct dose measurement can roughly halve the variability in the resulting concentrations. ⁶

2. Scan the phantom like a patient

Use the actual clinical whole-body acquisition and reconstruction protocol—matrix, iterations, filters, corrections, and time-of-flight settings. A phantom scanned with a special "QC" reconstruction does not test what patients actually receive. ⁴

3. Analyze consistently

Place background ROIs the same way each time, and use consistent, ideally automated, ROI or volume-of-interest analysis to remove observer variability. Automated ACR-phantom analysis has been shown to tighten measurement consistency and reduce operator dependence in SUV and cold-rod scoring. ^{4, 5}

4. Trend, don't just pass

Record background SUV, hot-cylinder SUVs and ratios, and cold-rod visibility over time. A value inside tolerance but trending steadily toward the limit is an early warning of calibration or detector drift—more useful than any single pass/fail. ⁴

Common pitfalls to avoid

• Fill and timing errors. An inaccurate activity assay or a clock mismatch corrupts the background SUV and masquerades as a scanner problem. ⁶
• Using a non-clinical protocol. QC that does not mirror the clinical protocol does not validate clinical SUV. ⁴
• Ignoring cross-calibration. The dose calibrator and the scanner must be cross-calibrated; a drift in either shifts SUV. ^{4, 8}
• Treating the ACR phantom as a substitute for acceptance testing. NEMA and AAPM TG-126 acceptance testing characterizes the system; the ACR phantom confirms ongoing accuracy. Both are needed. ^{7, 8}
• Skipping trend analysis. A single pass hides a drift that trending would catch. ⁴

Regulatory Considerations

PET/CT quality control sits at the intersection of radioactive-material regulation and imaging accreditation, and the phantom program is what makes the quantitative side defensible. The positron-emitting radiopharmaceuticals are byproduct or accelerator-produced material regulated under the medical-use framework of 10 CFR Part 35 (or the equivalent Agreement State program), while the CT subsystem is a registered X-ray machine regulated by the FDA and the states. Image-quality and quantitative accuracy, however, are governed by accreditation.

Key frameworks to reference:

• ACR PET/CT Accreditation Program and ACR PET phantom testing — require scanning the ACR phantom and an annual system performance evaluation performed by or under the direction of a qualified medical physicist, with defined acceptance for SUV, contrast, and uniformity. ⁴
• AAPM Report No. 126 (Task Group 126) — the manufacturer-independent PET/CT acceptance-testing and QA reference that defines the uniformity, correction-accuracy, and registration tolerances. ⁸
• NEMA NU 2 — the standardized performance-measurement methodology (sensitivity, spatial resolution, count-rate/NECR, and image quality) used at acceptance and to verify vendor specifications. ⁷
• Joint Commission imaging standards — require ongoing QC and an annual equipment performance evaluation by a qualified medical physicist for accredited organizations.

Because the states DRPS serves administer their own radioactive-material and X-ray programs (Florida, Maryland, Virginia, California, Nevada, Pennsylvania, New York, and New Jersey are NRC Agreement States, while Washington DC and Delaware are regulated directly by the NRC for byproduct material), a PET/CT program must reconcile its radioactive-material license, its CT registration, and its accreditation obligations. Documented phantom results, traceable calibrations, and a physicist's report are what make the quantitative program defensible during accreditation review. For the acceptance-testing companion to this QC, see NEMA NU 2 PET/CT performance testing.

Frequently Asked Questions (FAQs)

What is the ACR PET phantom and what does it test?

The ACR PET phantom is a cylindrical accreditation phantom containing hot cylinders of decreasing diameter, cold rods, and a uniform background compartment filled to a known activity concentration. When scanned with the laboratory's clinical protocol, it verifies that the PET/CT recovers accurate standardized uptake values (SUV), reproduces contrast in small structures, and produces a uniform image—so the quantitative numbers used clinically can be trusted.

Why does the ACR PET phantom background target an SUV of about 1.0?

The phantom is filled so the background activity concentration equals the injected activity divided by the phantom volume—the same ratio that defines SUV. By construction, a perfectly calibrated scanner should measure a background SUV of about 1.0. A measured background that drifts outside the ACR's acceptance window points to a calibration error in the scanner, the dose calibrator, or the clock synchronization between them.

How is the ACR PET phantom different from the NEMA NU 2 image-quality phantom?

The NEMA NU 2 image-quality phantom, with six fillable spheres from 10 to 37 mm, is used mainly at acceptance to measure contrast recovery and background variability against the manufacturer's specification. The ACR PET phantom is the accreditation and annual-QC tool: it uses hot cylinders and cold rods to confirm ongoing SUV accuracy and image quality with the laboratory's own clinical protocol. They are complementary, not interchangeable.

How often is PET/CT phantom QC required?

Daily QC (detector calibration, blank/normalization, and a well-counter or SUV calibration check) is run by the technologist; the ACR accreditation phantom is scanned by or under the supervision of a qualified medical physicist at least annually and at accreditation; and acceptance testing with NEMA and AAPM TG-126 methods is performed when a scanner is installed or after major service. Any calibration change or image-quality concern triggers additional testing.

What SUV accuracy does a PET/CT scanner need to meet?

AAPM Task Group 126 acceptance testing expects the accuracy of count-rate corrections, expressed as a background SUV, to fall within about 10% of the expected value of 1.0, image uniformity within roughly ±5%, and PET-to-CT image registration within about 1 mm. These tolerances keep SUV reliable enough for response assessment and multi-scanner comparison.

Why does SUV accuracy matter clinically?

SUV is used to characterize lesions, compare scans over time, and judge response to therapy using criteria such as PERCIST. If a scanner's calibration drifts, every SUV shifts with it, and a treatment response can be misclassified. Phantom QC is what proves the SUV a physician reads is accurate and comparable across scanners and across time.

Does a medical physicist have to perform the ACR PET phantom test?

Yes. The ACR PET/CT accreditation program requires that the phantom testing and the annual system performance evaluation be performed by, or under the direction of, a qualified medical physicist, who interprets the SUV, contrast, and uniformity results and directs corrective action before quantitative accuracy is affected.

Key Takeaways

• The ACR PET phantom validates quantitative accuracy end-to-end. It exercises the whole chain from dose calibrator to reconstruction and asks whether SUV comes out right. ⁴
• Background SUV should read about 1.0 by construction. A drift from 1.0 is a direct signature of a calibration error. ⁴
• Contrast recovery falls with object size. Partial-volume blurring means small structures never reach true SUV; CRC quantifies how much survives. ^{7, 10}
• Acceptance tolerances anchor the program. AAPM TG-126 expects correction accuracy within ~10% of SUV 1.0, uniformity within ~±5%, and registration within ~1 mm. ⁸
• ACR and NEMA/AAPM testing are complementary. NEMA and TG-126 characterize the system; the ACR phantom confirms ongoing accuracy with the clinical protocol. ^{7, 8}
• SUV accuracy is a clinical-decision issue. Calibration drift can misclassify treatment response, which is why phantom QC is not optional. ^{10, 11}

Conclusion

A PET/CT report often comes down to a number, and that number carries weight only if the scanner that produced it is proven accurate. The ACR PET phantom is the proof. By filling a phantom to a concentration that should read as an SUV of 1.0, scanning it the way patients are scanned, and confirming background SUV, contrast recovery, and cold-rod visibility, the medical physicist converts confidence into documented evidence.

A defensible PET/CT program layers this together: daily technologist QC, an annual ACR phantom evaluation by a qualified medical physicist, and NEMA/TG-126 acceptance testing at installation and after major service. That structure is what lets a laboratory stand behind every SUV it reports—across scanners, across sites, and across the months that separate a baseline scan from the follow-up that guides a patient's treatment.

How DRPS Can Help

Diagnostic Radiation Physics Services helps nuclear medicine and PET/CT facilities build quantitative quality programs that are accurate, documented, and accreditation-ready. This includes ACR PET phantom testing, annual PET/CT performance evaluations, NEMA and AAPM TG-126 acceptance testing, dose-calibrator and scanner cross-calibration, and ACR accreditation submission support—delivered through our PET/CT and nuclear medicine physics, accreditation support, and medical physicist consulting services.

DRPS supports facilities across our service locations, including Florida, Maryland, Virginia, Washington DC, California, Nevada, New York, Pennsylvania, New Jersey, and Delaware.

A strong PET/CT QC program does not just pass accreditation—it makes every SUV a number the care team can act on with confidence.

Related Resources

• PET/CT daily QC and calibration
• NEMA NU 2 PET/CT performance testing
• PET SUV quantification
• EARL PET SUV harmonization
• The PET partial-volume effect
• PET/CT and nuclear medicine physics
• Accreditation support
• Medical physicist consulting

References

1. Boellaard R, Delgado-Bolton R, Oyen WJG, et al. FDG PET/CT: EANM procedure guidelines for tumour imaging: version 2.0. European Journal of Nuclear Medicine and Molecular Imaging. 2015;42(2):328-354. doi:10.1007/s00259-014-2961-x. PubMed
2. Kinahan PE, Fletcher JW. Positron emission tomography-computed tomography standardized uptake values in clinical practice and assessing response to therapy. Seminars in Ultrasound, CT and MRI. 2010;31(6):496-505. doi:10.1053/j.sult.2010.10.001. PubMed
3. Wahl RL, Jacene H, Kasamon Y, Lodge MA. From RECIST to PERCIST: evolving considerations for PET response criteria in solid tumors. Journal of Nuclear Medicine. 2009;50(Suppl 1):122S-150S. doi:10.2967/jnumed.108.057307. PubMed
4. DiFilippo FP, Patel M, Patel S. Automated quantitative analysis of American College of Radiology PET phantom images. Journal of Nuclear Medicine Technology. 2019;47(3):249-254. doi:10.2967/jnmt.118.221317. PubMed
5. DiFilippo FP. Assessment of PET and SPECT phantom image quality through automated binary classification of cold rod arrays. Medical Physics. 2019;46(8):3451-3461. doi:10.1002/mp.13616. PubMed
6. Beattie BJ. Proposed changes to the ACR phantom filling procedure for more accurate and consistent activity concentrations. Journal of Applied Clinical Medical Physics. 2019;20(2):154-156. doi:10.1002/acm2.12531. PubMed
7. Kolthammer JA, Su KH, Grover A, Narayanan M, Jordan DW, Muzic RF. Performance evaluation of the Ingenuity TF PET/CT scanner with a focus on high count-rate conditions. Physics in Medicine and Biology. 2014;59(14):3843-3859. doi:10.1088/0031-9155/59/14/3843. PubMed
8. AlMazrou RY, Alanazi SF, Alzaid MH, Al-Fakhranee RS, Ding S, Mawlawi OR. Performance characteristics of the 5-ring GE Discovery MI PET/CT scanner using AAPM TG-126 report. Journal of Applied Clinical Medical Physics. 2024;25(4):e14315. doi:10.1002/acm2.14315. PubMed
9. Aide N, Lasnon C, Veit-Haibach P, Sera T, Sattler B, Boellaard R. EANM/EARL harmonization strategies in PET quantification: from daily practice to multicentre oncological studies. European Journal of Nuclear Medicine and Molecular Imaging. 2017;44(Suppl 1):17-31. doi:10.1007/s00259-017-3740-2. PubMed
10. Monsef A, Ay MR, Sheikhzadeh P, Geramifar P, Rahmim A, Ghafarian P. Harmonization based on quantitative analysis of standardized uptake value variations across PET/CT scanners: a multicenter phantom study. Nuclear Medicine Communications. 2022;43(9):1004-1014. doi:10.1097/MNM.0000000000001598. PubMed
11. Mattoli MV, Calcagni ML, Taralli S, Indovina L, Spottiswoode BS, Giordano A. How often do we fail to classify the treatment response with [18F]FDG PET/CT acquired on different scanners? Insights from a phantom study. Molecular Imaging and Biology. 2019;21(6):1210-1219. doi:10.1007/s11307-019-01342-5. PubMed