MRI Acceptance Testing and Commissioning

Introduction

MRI acceptance testing and commissioning is the comprehensive, one-time performance evaluation a qualified medical physicist performs when a magnet is newly installed, relocated, or substantially upgraded — before the first patient is scanned. It confirms the system meets its purchase specifications and the applicable standards, documents the safety systems, and establishes the quantitative baselines that every later quality-control (QC) test is measured against.

Acceptance testing is easy to confuse with routine QC, but they answer different questions. Routine QC asks, "Is the scanner still performing the way it did at baseline?" Acceptance testing asks the prior question: "What is the baseline, and does this system actually meet what we paid for and what the standards require?" A facility that skips a rigorous acceptance evaluation has no defensible reference point — every weekly phantom scan afterward is compared against a number nobody independently verified.

This guide walks through what acceptance testing covers, the standards that govern it, the quantitative tests and their representative criteria, a worked signal-to-noise example, and how the process connects to accreditation and the annual physicist evaluation. DRPS performs MRI acceptance and annual performance testing as part of its MRI physics testing and medical physics consulting services across Florida, Maryland, Virginia, Washington DC, California, Nevada, Pennsylvania, New York, New Jersey, and Delaware.

Topic Explanation

What is MRI acceptance testing and commissioning?

Acceptance testing is the formal demonstration that an installed MRI system meets its contracted specifications and applicable performance standards; commissioning is the broader process of getting that system ready for clinical use, including baseline establishment, protocol verification, and safety sign-off. In practice the two overlap, and a physicist's acceptance report is the central document of commissioning.

The scope is deliberately wide. Acceptance testing is not a single phantom scan — it is a system-level evaluation that spans the magnet, gradients, radiofrequency (RF) subsystem, receiver chain, the installed coils, the reconstruction software, and the room itself. AAPM Report No. 100, the primary U.S. guidance on the subject, was written to help the medical physicist "define an acceptance test strategy and quality assurance procedures" for a facility, and it explicitly notes that some acceptance criteria are set tighter than the action criteria in the ACR program, because acceptance testing should establish the optimal operating characteristics of the system. ¹

A practical acceptance evaluation starts with a short set of facility-level questions:

Does the magnet meet its homogeneity and stability specification over the specified diameter of spherical volume (DSV)?
Is the RF shielding (the room's "Faraday cage") intact, so external RF does not corrupt images?
Do all delivered coils meet their signal-to-noise and uniformity specifications?
Is geometric accuracy adequate across the field of view the clinical protocols actually use?
Are the safety systems — quench vent, oxygen monitor, door interlock, emergency stop, and cryogen alarms — present, labeled, and functional?
Have reproducible baselines been recorded for every parameter the ongoing QC program will track?

How acceptance testing differs from routine QC

The routine program that most technologists know — the weekly or daily ACR phantom scan and the annual physicist evaluation — is covered in our companion guide to MRI ACR phantom QC and its seven core tests. That program is intentionally abbreviated and repeatable. Acceptance testing is the opposite: a deep, one-time evaluation that uses a wider instrument set and tighter criteria, and that produces the reference values the routine program then defends. ^{1, 2}

Attribute	Acceptance testing / commissioning	Routine quality control
When performed	Once, at install, relocation, or major upgrade	Repeated: daily/weekly (technologist) and annually (physicist)
Scope	System-wide: magnet, gradients, RF, coils, shielding, safety	Abbreviated phantom tests plus annual physicist review
Criteria source	Purchase specification, AAPM Report No. 100, NEMA MS series, ACR	ACR action limits and facility baselines
Typical strictness	Often tighter than ACR action limits ¹	ACR action limits
Primary output	Baselines, specification compliance, safety sign-off	Trend against baseline; pass/fail against action limits
Who performs	Qualified medical physicist (independent of vendor)	Technologist (routine) and physicist (annual)

Key Technical Principles

Magnetic field homogeneity

A uniform static field $B_{0}$ is the foundation of MR image quality; inhomogeneity causes geometric distortion, failed fat suppression, and signal loss. Homogeneity is quantified as a peak-to-peak or root-mean-square deviation over a defined DSV, expressed in parts per million (ppm):

Δ B_{ppm} = \frac{B _{m a x} - B _{m i n}}{B _{0}} \times 1 0^{6}

Acceptance testing verifies homogeneity against the magnet specification using a spectral-peak (phase-difference) or field-mapping method, typically over nested spherical volumes (for example 20, 30, and 40 cm DSV). The measured values must meet the vendor specification after shimming, and they become the baseline for detecting later drift. ¹

Center frequency and system stability

The center (Larmor) frequency $f_{0}$ is set by the field and the gyromagnetic ratio $γ$ :

f_{0} = \frac{γ}{2 π} B_{0}

For hydrogen, $γ /2 π \approx 42.58 MHz/T$ , so a 1.5 T system resonates near 63.9 MHz and a 3.0 T system near 127.7 MHz. Acceptance testing records the center frequency and confirms short-term stability; a drifting center frequency is an early sign of field or thermal instability and can degrade fat saturation and spectroscopy. ¹

Signal-to-noise ratio and the two-image method

Signal-to-noise ratio (SNR) is the single most sensitive indicator of overall system and coil performance. The NEMA MS 1 approach measures SNR from two identical, consecutive acquisitions: the mean signal is taken from a region of interest (ROI) in the uniform phantom, and the noise is taken from the standard deviation of the subtraction (difference) image, which removes fixed structure. ⁴

SNR = 2 \frac{S ˉ}{σ _{diff}}

where $\overset{ˉ}{S}$ is the mean ROI signal in one image and $σ_{diff}$ is the standard deviation of the pixel values in the difference image over the same ROI. The factor of $2$ corrects for the fact that subtracting two images adds their noise in quadrature. Acceptance testing records a baseline SNR for each coil; because SNR scales with field strength, coil loading, and sequence parameters, the baseline must be tied to a fixed, documented protocol so later comparisons are valid.

Worked SNR example

Suppose a head-coil acceptance series gives a mean ROI signal of $\overset{ˉ}{S} = 1500$ (arbitrary units) and a difference-image noise of $σ_{diff} = 30$ over the same ROI. The baseline SNR is:

SNR = 2 \times \frac{1500}{30} \approx 1.414 \times 50 \approx 71

This value is only meaningful relative to the fixed protocol it was acquired with — the same sequence, bandwidth, field of view, slice thickness, and coil. That is exactly why acceptance testing documents the acquisition conditions alongside the number: an SNR of 71 means nothing to next year's physicist unless the baseline protocol is recorded with it. The numbers here are illustrative of the method, not a specification for any particular scanner.

Image uniformity and ghosting

Image intensity uniformity describes how evenly a homogeneous phantom appears across the field of view. The percent integral uniformity (PIU) used in the ACR program is:

PIU = 100 (1 - \frac{S _{m a x} - S _{m i n}}{S _{m a x} + S _{m i n}})

where $S_{m a x}$ and $S_{m i n}$ are the highest and lowest mean signals in small ROIs placed within a large ROI over the phantom. NEMA MS 3 describes complementary uniformity metrics. ^{3, 5} The ACR acceptance values are PIU greater than or equal to 87.5% for systems below 3 T and greater than or equal to 82.0% at 3 T and above. ³

Ghosting — a faint, displaced replica of the phantom in the phase-encode direction — is quantified by a ghosting ratio built from ROIs placed outside the phantom (top, bottom, left, right) and one inside:

GR = \frac{( top + bottom ) - ( left + right )}{2 \times ROI _{center}}

The ACR acceptance value for percent signal ghosting is a ratio at or below 0.025. ³

RF shielding integrity

The scan room is an RF-shielded enclosure that keeps external radio signals out of the receiver bandwidth. A defect — a gap in the shielding, an ungrounded penetration, or a compromised door seal — shows up as discrete "zipper" or broadband noise in images. Acceptance testing includes an RF shielding evaluation, usually performed by the shielding vendor or physicist, confirming the enclosure meets its attenuation specification before the system is accepted. AAPM Report No. 100 treats the physicist's involvement in RF shielding as part of acceptance, from purchase through construction. ¹

Clinical Impact

A thorough acceptance evaluation protects image quality, patient safety, and the facility's investment. Three consequences follow directly from doing it well — or badly.

First, baselines. Every later QC decision — "has SNR dropped?", "is uniformity degrading?", "did that service event change geometric accuracy?" — is only as good as the baseline it is compared against. If the baseline was never independently measured, the facility cannot distinguish a real failure from normal variation, and problems are caught late, often after image quality has visibly degraded. ^{1, 2}

Second, specification compliance. MRI systems are major capital purchases, and acceptance testing is the facility's leverage to confirm the delivered system meets what was contracted — homogeneity over the specified DSV, SNR for each coil, gradient performance, and software features. Discrepancies identified at acceptance are the vendor's responsibility to correct; the same discrepancies found a year later are the facility's problem.

Third, geometric accuracy for advanced use. Distortion grows with distance from isocenter because gradient nonlinearity and field inhomogeneity worsen toward the edge of the field of view. For routine diagnostic imaging a small distortion is tolerable, but for MR-guided radiation therapy planning, stereotactic work, or large field-of-view imaging, geometric fidelity across the whole volume must be characterized at acceptance, sometimes with a large field-of-view distortion phantom beyond the standard ACR phantom. Multi-site experience with the ACR phantom shows geometric accuracy can and does fail in practice, reinforcing the need to measure it rather than assume it. ⁸

Practical Optimization Tips

A defensible acceptance evaluation follows a consistent workflow.

1. Review the purchase specification first

The contract defines what "acceptable" means for this system: homogeneity over a stated DSV, per-coil SNR, gradient strength and slew rate, and included software. Acceptance tests should map directly onto those line items so compliance (or a deficiency) is unambiguous.

2. Coordinate with vendor installation, but stay independent

The vendor calibrates and tunes the system; the physicist independently verifies it. These are complementary, not redundant. The physicist's report should stand on its own, with instruments, methods, and measured values documented.

3. Run the ACR phantom tests, then go beyond them

Acquire the ACR phantom series so baselines align with the accreditation program, then add the system-level tests the ACR program does not fully cover: field homogeneity, center frequency and stability, fringe-field verification, transmitter gain/reference calibration, eddy-current and gradient checks, EPI (echo-planar) ghosting and spike-noise screening, and RF shielding. ^{1, 3}

4. Document baselines with their acquisition conditions

A baseline SNR, uniformity, or ghosting value is meaningless without the exact protocol that produced it. Record sequence, bandwidth, field of view, matrix, slice, coil, and phantom so the annual and routine programs can reproduce it.

5. Verify safety systems explicitly

Confirm and document the quench vent path, Zone IV oxygen monitor, door interlock, emergency-stop/quench button, cryogen-level and pressure alarms, acoustic-noise and specific-absorption-rate (SAR) display behavior, and MR-conditional labeling in the environment. Static-field and SAR limits follow IEC 60601-2-33. ⁶

Common pitfalls to avoid

Treating acceptance as "a phantom scan." It is a system-level evaluation, not a single image-quality test.
Accepting vendor numbers without independent measurement. The physicist's report is the facility's evidence of compliance.
Recording baselines without protocols. An un-reproducible baseline is useless to the ongoing program.
Ignoring geometric accuracy at the edge of the field of view. Distortion is largest exactly where advanced applications need fidelity.
Skipping re-baselining after major service. A gradient swap, coil replacement, quench-and-ramp, or major upgrade changes performance and demands a focused re-evaluation.

Regulatory Considerations

MRI is non-ionizing, so it is not regulated under the NRC or state radiation-machine x-ray programs — but acceptance and annual performance testing are still required by accreditation and hospital-standard frameworks. This is an important distinction: unlike CT or fluoroscopy, an MRI system's performance obligations come from accreditation and facility standards rather than from a radiation-control statute.

Key frameworks:

ACR MRI Accreditation Program. Accredited facilities must pass a phantom-based image-quality review and undergo an annual MRI equipment performance evaluation by a qualified medical physicist or MR scientist. The ACR MRI Quality Control Manual (2015) and the Large and Medium Phantom Test Guidance (2022) define the tests and acceptance criteria. ^{2, 3, 11}
AAPM Report No. 100. The U.S. medical-physics community's guidance on acceptance testing and QA strategy, including criteria that may exceed ACR action limits. ¹
NEMA MS measurement standards. Standardized definitions for SNR (MS 1), image uniformity (MS 3), slice thickness (MS 5), and related parameters, so measurements are comparable across systems and vendors. ^{4, 5, 7}
IEC 60601-2-33. The international safety standard for MR equipment, including static-field access and SAR operating modes that acceptance testing confirms are correctly implemented. ⁶

Because MRI has no MQSA-style federal quality mandate, the accreditation and annual-physicist requirements effectively are the compliance floor for most U.S. imaging facilities. For the safety side of commissioning — radiofrequency heating and the specific-absorption-rate limits a physicist confirms at acceptance — see our guide to MRI SAR and RF safety. Facilities adding a magnet should connect acceptance testing to accreditation support and ongoing MRI physics testing.

Frequently Asked Questions (FAQs)

What is MRI acceptance testing?

MRI acceptance testing is the comprehensive performance evaluation a qualified medical physicist performs when a scanner is newly installed, relocated, or substantially upgraded. It confirms the system meets purchase specifications and applicable standards, documents safety features, and establishes the quantitative baselines used to judge every later quality-control measurement.

How is acceptance testing different from routine MRI QC?

Acceptance testing is a one-time, in-depth evaluation at installation that often applies criteria stricter than ongoing quality control, while routine QC is a repeated, abbreviated set of checks (daily technologist tests and the annual physicist evaluation) that compares current performance against the baselines acceptance testing established.

Who performs MRI acceptance testing?

A qualified or board-certified medical physicist performs or directs MRI acceptance testing, usually alongside the vendor's installation and calibration. The physicist's report is independent of the vendor and documents whether the system meets specifications before clinical use.

What tests are included in MRI acceptance testing?

Typical tests include magnetic field homogeneity, center (Larmor) frequency, signal-to-noise ratio, image uniformity, geometric accuracy, slice thickness and position, high-contrast resolution, low-contrast detectability, ghosting and artifacts, RF shielding integrity, gradient performance, and verification of safety systems.

Does acceptance testing use the ACR phantom?

It usually includes the ACR phantom tests so that baselines align with the accreditation program, but acceptance testing is broader. It also evaluates system-level parameters such as field homogeneity, RF shielding, transmitter calibration, and fringe field that the routine ACR phantom program does not fully cover.

What acceptance criteria are used for MRI?

Criteria come from the equipment purchase specification, the ACR phantom test guidance, the NEMA MS measurement standards, and AAPM Report No. 100. AAPM Report No. 100 notes that acceptance criteria are often set tighter than ACR action limits because acceptance testing should establish the system's optimal operating characteristics.

When should MRI acceptance testing be repeated?

Repeat a focused acceptance or re-baseline evaluation after a magnet relocation, a gradient or RF coil replacement, a major software or hardware upgrade, a quench and ramp, or any service event that could change image quality or safety performance.

Key Takeaways

Acceptance testing establishes the baseline; routine QC defends it. Without an independently measured baseline, every later QC comparison is guesswork.
It is a system-level evaluation, not a phantom scan. Magnet, gradients, RF, coils, shielding, and safety systems are all in scope.
Criteria are often tighter than ACR action limits. AAPM Report No. 100 sets acceptance criteria to capture the system's optimal performance, not just its minimum acceptable performance. ¹
Representative ACR acceptance values include geometric accuracy within about ±2 mm, slice thickness 5.0 mm ± 0.7 mm, PIU at or above 87.5% (below 3 T) or 82.0% (3 T and above), and a ghosting ratio at or below 0.025. ³
Document baselines with their protocols. A number without its acquisition conditions cannot be reproduced or trended.
Re-baseline after major service. Gradient or coil swaps, quenches, and major upgrades change performance and require a focused re-evaluation.

Conclusion

MRI acceptance testing and commissioning is the moment a facility converts a capital purchase into a clinically defensible imaging system. It verifies that the delivered magnet meets its specification, confirms that the safety systems work, and — most durably — fixes the quantitative baselines that the entire quality-control program will rely on for the life of the scanner.

Done rigorously, acceptance testing pays for itself: it catches specification shortfalls while they are still the vendor's responsibility, it gives the annual physicist a real reference instead of an assumption, and it documents the system in enough detail that a service event years later can be evaluated against a known-good state. A magnet that was never properly commissioned is a magnet whose performance nobody can actually vouch for.

How DRPS Can Help

Diagnostic Radiation Physics Services performs MRI acceptance testing, commissioning, and annual equipment performance evaluations for imaging facilities, hospitals, and outpatient centers. Our board-certified physicists verify systems against purchase specifications, run the ACR phantom and system-level tests, document reproducible baselines, confirm safety-system function, and prepare the reports accreditation and administration require.

DRPS supports facilities across our service locations, including Florida, Maryland, Virginia, Washington DC, California, Nevada, New York, Pennsylvania, New Jersey, and Delaware. Whether you are installing your first magnet, relocating a system, or re-baselining after a major upgrade, we make the acceptance process defensible and the baselines trustworthy.

Related Resources

References

American Association of Physicists in Medicine. Acceptance Testing and Quality Assurance Procedures for Magnetic Resonance Imaging Facilities. AAPM Report No. 100. College Park, MD: AAPM; 2010. aapm.org
American College of Radiology. MRI Quality Control Manual. Reston, VA: ACR; 2015. acraccreditation.org
American College of Radiology. Large and Medium Phantom Test Guidance for the ACR MRI Accreditation Program. Reston, VA: ACR; 2022. acr.org
National Electrical Manufacturers Association. Determination of Signal-to-Noise Ratio (SNR) in Diagnostic Magnetic Resonance Imaging. NEMA Standards Publication MS 1-2008 (R2020). Rosslyn, VA: NEMA. nema.org
National Electrical Manufacturers Association. Determination of Image Uniformity in Diagnostic Magnetic Resonance Images. NEMA Standards Publication MS 3-2008 (R2020). Rosslyn, VA: NEMA. nema.org
International Electrotechnical Commission. Medical electrical equipment — Part 2-33: Particular requirements for the basic safety and essential performance of magnetic resonance equipment for medical diagnosis. IEC 60601-2-33. Geneva: IEC. iec.ch
National Electrical Manufacturers Association. Determination of Slice Thickness in Diagnostic Magnetic Resonance Imaging. NEMA Standards Publication MS 5-2018. Rosslyn, VA: NEMA. nema.org
Ihalainen TM, Lönnroth NT, Peltonen JI, et al. MRI quality assurance using the ACR phantom in a multi-unit imaging center. Acta Oncol. 2011;50(6):966-972. doi:10.3109/0284186X.2011.582515. PubMed
Panych LP, Chiou JG, Qin L, et al. On replacing the manual measurement of ACR phantom images performed by MRI technologists with an automated measurement approach. J Magn Reson Imaging. 2016;43(4):843-852. doi:10.1002/jmri.25052. PubMed
Alaya IB, Mars M. Automatic analysis of ACR phantom images in MRI. Curr Med Imaging. 2020;16(7):892-901. doi:10.2174/1573405615666190903145343. PubMed
American College of Radiology. MRI Accreditation Program Requirements. Reston, VA: ACR. acraccreditation.org