How often should a Jaszczak phantom SPECT test be performed?

Multipurpose SPECT phantom testing is typically performed quarterly as part of a routine quality-control program and at acceptance, consistent with AAPM Task Group 177 recommendations and ACR accreditation requirements. Daily and weekly gamma-camera tests such as uniformity and energy peaking continue on their own schedules.

Why are the spheres and rods in the phantom cold?

The rods and spheres are solid, non-radioactive plastic suspended in a uniform background of radioactive solution, so they appear as cold (low-count) defects. Cold contrast objects test the system's ability to detect a lesion-like reduction in counts against a warm background, which is the clinically relevant task for many SPECT studies.

What causes ring artifacts in a Jaszczak phantom SPECT image?

Ring or bullseye artifacts in a reconstructed uniform region are the classic sign of non-uniform detector response combined with the rotational geometry of SPECT. A small non-uniformity that is tolerable in planar imaging is amplified into a ring by filtered backprojection, which is why high-count uniformity correction floods and center-of-rotation calibration are prerequisites for a good phantom result.

Does a medical physicist need to be involved in SPECT phantom QC?

Yes. A qualified or board-certified medical physicist establishes the acquisition and reconstruction protocol, sets tolerances, interprets the phantom images, and performs acceptance and annual evaluations. ACR accreditation and AAPM Task Group 177 both frame SPECT phantom testing within a physicist-directed quality-control program.

SPECT Jaszczak Phantom QC: Resolution & Contrast

Q: What does the Jaszczak SPECT phantom measure?

A Jaszczak-style SPECT phantom measures three tomographic performance characteristics in a single acquisition: reconstructed (tomographic) spatial resolution from a sector of cold rods, contrast detectability from a set of cold solid spheres, and tomographic uniformity from a large uniform region. It reveals problems that planar tests cannot, such as center-of-rotation errors and reconstruction artifacts.

Q: How is cold-sphere contrast quantified?

Contrast is quantified by comparing the mean counts in a region of interest over a cold sphere with the mean counts in the surrounding warm background, using the expression C = (background − sphere) / background. The smallest sphere that is reliably visible, or that meets a contrast-to-noise threshold near the Rose criterion, characterizes contrast detectability.

A gamma camera can pass every planar quality-control test and still reconstruct poor tomographic images. The Jaszczak-style SPECT phantom — cold rods, cold spheres, and a uniform region in one water-filled cylinder — is the single acquisition that exposes reconstructed spatial resolution, contrast detectability, and tomographic uniformity together, revealing center-of-rotation and reconstruction problems that planar tests never see.¹²³ That is why it anchors both routine SPECT quality control and ACR accreditation.

This guide covers what the phantom contains and why, how center-of-rotation and uniformity errors turn into ring artifacts, the physics of reconstructed resolution and cold-sphere contrast with worked math, how to fill and acquire the phantom correctly, how the results are quantified, and where the test fits in a physicist-directed QC and accreditation program.¹²⁴

Introduction

SPECT adds a dimension to planar imaging, and it adds failure modes with it. In single-photon emission computed tomography the camera rotates around the patient, and projections are reconstructed into transaxial slices. Reconstruction is unforgiving: a small detector non-uniformity, a fraction-of-a-pixel center-of-rotation offset, or a poorly chosen filter can each destroy tomographic image quality even when planar uniformity and resolution look acceptable.¹³⁴

The multipurpose SPECT phantom introduced by Jaszczak and colleagues solved a practical problem — how to verify the whole tomographic chain with one physical object. A single cylinder holds three test regions: a set of cold rod sectors for reconstructed spatial resolution, a set of cold solid spheres for contrast detectability, and a large uniform volume for tomographic uniformity. Fill it with a well-mixed radioactive solution, acquire a full SPECT study, reconstruct it exactly as clinical studies are reconstructed, and the phantom reports on resolution, contrast, uniformity, center of rotation, and reconstruction artifacts all at once.¹²³

This test sits at the center of both routine SPECT QC and accreditation. The ACR Nuclear Medicine and PET Accreditation Program requires phantom imaging of a Jaszczak-derived design, and AAPM Task Group 177 places multipurpose phantom evaluation in the recommended acceptance and periodic survey program for gamma camera, SPECT, and SPECT/CT systems.²⁷

Topic Explanation

What the phantom contains

A Jaszczak-style phantom is a water-fillable acrylic cylinder. The ACR-approved version, widely used for accreditation, has an internal radius of about 10.8 cm and an internal length of about 20 cm, and it contains three functional regions:⁷

Cold rod sectors — six pie-shaped sectors of solid acrylic rods, with rod diameters of 4.8, 6.4, 7.9, 9.5, 11.1, and 12.7 mm. The rods appear cold against the warm background and test reconstructed spatial resolution: the smallest sector whose rods can be resolved characterizes tomographic resolution.⁷
Cold solid spheres — six solid acrylic spheres with diameters of 9.5, 12.7, 15.9, 19.1, 25.4, and 31.8 mm. They appear as cold defects and test contrast detectability: the smallest sphere reliably seen characterizes low-contrast performance.⁷
Uniform region — a large open volume of radioactive solution used to assess tomographic (reconstructed) uniformity and to reveal ring artifacts.¹²

The design is deliberate. Rods probe high-spatial-frequency performance, spheres probe low-contrast detection, and the uniform region probes the systematic errors that reconstruction amplifies.¹³

Why planar QC is not enough

Daily and weekly gamma-camera tests — energy peaking, intrinsic and extrinsic flood uniformity, and bar-phantom resolution — verify the detector as a projection imager. SPECT adds three things planar tests cannot check:³⁴

Center of rotation (COR). The reconstruction assumes the axis of rotation projects to a known column in every view. A COR offset of even a fraction of a pixel blurs and splits point sources into "tuning-fork" or doughnut shapes in the reconstruction.¹⁴
Rotational uniformity. A detector non-uniformity that is invisible in a single planar flood becomes a concentric ring in the reconstructed uniform region because backprojection smears it around the rotation axis.¹³
Reconstruction and filter behavior. The choice of filtered backprojection versus iterative reconstruction, the reconstruction filter, and attenuation and scatter correction all change resolution, contrast, and noise — and the phantom shows their net effect.³⁵

Key Technical Principles

Reconstructed spatial resolution

SPECT spatial resolution is dominated not by the detector but by the collimator, and it degrades with distance from the collimator face. For a parallel-hole collimator with hole diameter $d$ , effective hole length $L_{e}$ , and source-to-collimator distance $b$ , the geometric collimator resolution is:³⁴

R_{coll} = d \cdot \frac{L _{e} + b}{L _{e}}

The system spatial resolution combines the collimator and intrinsic (detector) resolutions in quadrature:

R_{sys} = R_{coll}^{2} + R_{int}^{2}

Because $R_{coll}$ grows linearly with source distance $b$ , resolution is always worse at the center of the phantom (and the patient) than at the surface — one reason the phantom must be centered in the field of view and the radius of rotation kept as small as safely possible.³⁴ The reconstruction filter further shapes the final resolution: a smoother filter suppresses noise but widens the reconstructed point-spread function.³⁵

Cold-sphere contrast and detectability

Contrast for a cold sphere is defined by comparing counts in a region of interest over the sphere with counts in the surrounding warm background:⁶

C = \frac{N ˉ _{bkg} - N ˉ _{sphere}}{N ˉ _{bkg}}

A perfect cold sphere with no partial-volume blurring would give $C = 1$ ; real spheres give less, and smaller spheres give less still because the finite reconstructed resolution "fills in" cold counts (the partial-volume effect). Detectability depends on both contrast and noise, expressed as a contrast-to-noise ratio:

CNR = \frac{N ˉ _{bkg} - N ˉ _{sphere}}{σ _{bkg}}

By the Rose criterion, an object is reliably detected when its CNR exceeds roughly 3 to 5; automated ACR-phantom analyses use CNR thresholds near this range to define the smallest visible sphere, and they agree well with expert visual reads.⁶ This is why phantom fill activity and acquisition counts matter: too few counts inflate $σ_{bkg}$ and hide spheres that adequate statistics would reveal.

Tomographic uniformity and ring artifacts

Reconstructed uniformity is quantified over the uniform region much as planar uniformity is, using an integral-uniformity expression:¹³

IU = \frac{N _{m a x} - N _{m i n}}{N _{m a x} + N _{m i n}} \times 100%

The characteristic SPECT failure is the ring (bullseye) artifact: a concentric ring centered on the axis of rotation. It arises when detector response is even slightly non-uniform, because backprojection distributes that local error around the rotation axis. The correction is preventive — a high-count uniformity-correction flood (typically tens of millions of counts) and current COR calibration must be in place before the phantom is acquired, or the ring will appear regardless of how well the phantom itself is prepared.¹³⁴

A worked resolution and contrast example

Consider a low-energy high-resolution (LEHR) collimator with hole diameter $d = 1.5$ mm and effective hole length $L_{e} = 35$ mm, imaging technetium-99m. Right at the collimator face ( $b = 0$ ):

R_{coll} = 1.5 \cdot \frac{35 + 0}{35} = 1.5 mm

At the center of the ACR phantom, roughly 10 cm from the collimator ( $b = 100$ mm):

R_{coll} = 1.5 \cdot \frac{35 + 100}{35} \approx 5.8 mm

Combining that with an intrinsic resolution of $R_{int} = 3.8$ mm gives a system resolution at depth of:

R_{sys} = 5. 8^{2} + 3. 8^{2} \approx 6.9 mm

This is exactly why the 4.8 mm and 6.4 mm rod sectors are the demanding ones: at phantom-center depths they sit near the limit of what the system can resolve, so failing to resolve them may reflect geometry and reconstruction rather than a broken detector.³⁴

For contrast, suppose a background region of interest averages $\overset{ˉ}{N}_{bkg} = 1000$ counts, so by Poisson statistics $σ_{bkg} = 1000 \approx 32$ counts, and a 12.7 mm cold sphere region averages $\overset{ˉ}{N}_{sphere} = 700$ counts. Then:

C = \frac{1000 - 700}{1000} = 0.30, CNR = \frac{1000 - 700}{32} \approx 9.4

A CNR near 9 is comfortably above the Rose criterion, so this sphere should be clearly visible. A smaller sphere with only $C = 0.10$ would give CNR ≈ 3.1 — right at the detectability threshold — which is why acquired counts, not just contrast, decide the smallest sphere a system can show.⁶

The Jaszczak phantom parameter map

Phantom region	Feature sizes	Performance characteristic	Typical failure it reveals
Cold rod sectors	4.8–12.7 mm rods	Reconstructed spatial resolution	COR error, over-smoothing filter
Cold solid spheres	9.5–31.8 mm spheres	Contrast detectability	Low counts, poor scatter/attenuation correction
Uniform region	Full cylinder volume	Tomographic uniformity	Ring/bullseye artifact from flood non-uniformity

All three are read from one reconstructed data set, which is what makes the phantom so efficient.¹²⁷

Clinical Impact

Tomographic performance is what determines whether a small perfusion defect, a subtle bone lesion, or a low-uptake focus is seen or missed. The phantom's three regions map directly onto clinical tasks: rod resolution corresponds to resolving adjacent structures, sphere contrast corresponds to detecting a lesion against background uptake, and uniformity corresponds to not mistaking a reconstruction ring for pathology.²³

A ring artifact is the most consequential of these because it can mimic disease — a bullseye across a myocardial perfusion or bone SPECT study can be read as a real defect. Because reconstruction amplifies uniformity errors, the phantom serves as an early warning: a ring that appears in the uniform region signals that the daily flood and correction map need attention before patient images are affected.¹³

On modern SPECT/CT systems the phantom result also underpins quantitative imaging. When a study reports absolute activity concentration or a standardized uptake value, the reconstructed resolution and uniformity that the phantom verifies feed directly into partial-volume behavior and recovery coefficients, so a system that fails phantom resolution will also bias small-lesion quantification. The same acquisition therefore serves both the qualitative reader, who needs artifact-free images, and the quantitative program, which needs a stable, characterized system before any activity measurement can be trusted.²³

Reconstruction choices also carry clinical weight. Studies comparing accreditation-style filtered backprojection with clinical iterative (OSEM) reconstructions on the same phantom show measurable differences in how many rods and spheres are visible, underscoring that a facility should evaluate the phantom under the reconstruction it actually uses clinically, not only under the settings an accrediting body suggests.⁵ Automated, reader-independent phantom analysis further improves consistency by removing subjective variability from the pass/fail decision.⁵⁶

Practical Optimization Tips

Fill and prepare the phantom carefully

Most phantom failures are preparation failures, not equipment failures:²³

Eliminate air bubbles. Trapped air under the lid or around the inserts creates false cold defects and non-uniformities. Fill completely and invert to clear bubbles.
Mix the activity thoroughly. Incomplete mixing produces activity gradients that read as uniformity failures. Add radionuclide to a partially filled phantom, then top off and agitate.
Use an appropriate activity and count density. Follow the accreditation or task-group guidance so that total counts are high enough to keep sphere detectability count-limited by design, not by under-acquisition.²⁷
Center the phantom in the field of view and minimize the radius of rotation, since collimator resolution degrades with distance.³⁴

Get the prerequisites right first

The phantom cannot compensate for an uncalibrated system. Before acquiring:¹⁴

Confirm the photopeak energy window is centered.
Apply a current, high-count uniformity-correction flood.
Verify center-of-rotation calibration for each collimator and acquisition mode.

Acquire the phantom with the same matrix, collimator, orbit, and number of views used clinically, and reconstruct it with the clinical filter and corrections so the result reflects real practice.²³

Analyze quantitatively and trend over time

Visual pass/fail of "smallest rods/spheres visible" is the traditional read, but quantitative analysis is more reproducible. Computing rod-sector modulation, sphere CNR, and integral uniformity — and trending them across quarters — detects gradual drift long before it becomes a visual failure. Automated analysis packages for the ACR phantom have been shown to correlate well with expert readers while improving precision and removing inter-observer variability.⁵⁶

Regulatory Considerations

SPECT systems are imaging devices, so their performance is governed primarily by accreditation and professional standards rather than by NRC materials rules — but the radioactive material used to fill the phantom falls under the facility's radioactive-material program.

Accreditation. The ACR Nuclear Medicine and PET Accreditation Program requires phantom imaging using a Jaszczak-derived (ACR-approved) design, with submitted images evaluated for reconstructed resolution, contrast, and uniformity.⁷
Professional QC standards. AAPM Report No. 22 established SPECT acceptance testing and quality control, and AAPM Report No. 177 (Task Group 177) provides current acceptance-testing and annual-survey recommendations for gamma camera, SPECT, and SPECT/CT systems.¹² The IAEA's Quality Assurance for SPECT Systems (Human Health Series No. 6) gives detailed acceptance, reference, and routine test procedures used internationally.³
Radioactive material. The technetium-99m or other radionuclide used to fill the phantom is byproduct material handled under 10 CFR Part 35 (or the equivalent Agreement State program) and 10 CFR Part 20. In Florida, medical use of radioactive material is administered under Florida Administrative Code Chapter 64E-5; DRPS also serves Maryland, Virginia, Washington DC, California, Nevada, Pennsylvania, New York, New Jersey, and Delaware, where NRC or Agreement State authorities impose parallel requirements. Always confirm requirements with the authority having jurisdiction.

Frequently Asked Questions (FAQs)

What does the Jaszczak SPECT phantom measure?

It measures reconstructed spatial resolution (from cold rod sectors), contrast detectability (from cold solid spheres), and tomographic uniformity (from a uniform region) in a single SPECT acquisition, while also revealing center-of-rotation errors and ring artifacts.¹²⁷

How often should the phantom test be performed?

Multipurpose SPECT phantom testing is typically quarterly and at acceptance, consistent with AAPM Task Group 177 and ACR accreditation, while daily and weekly gamma-camera tests continue on their own schedules.²⁷

Why are the rods and spheres cold rather than hot?

They are solid, non-radioactive plastic in a warm background, so they appear as cold (low-count) defects — the clinically relevant task of detecting a reduction in counts against surrounding uptake.⁶⁷

What causes a ring artifact in the reconstructed image?

Detector non-uniformity combined with rotational geometry: backprojection smears a small flood non-uniformity into a concentric ring. A high-count uniformity-correction flood and correct COR calibration prevent it.¹³

How is cold-sphere contrast quantified?

By comparing mean counts over a sphere with the warm background, $C = (\overset{ˉ}{N}_{bkg} - \overset{ˉ}{N}_{sphere}) / \overset{ˉ}{N}_{bkg}$ ; detectability is judged by a contrast-to-noise ratio near the Rose criterion of about 3 to 5.⁶

Should the phantom be reconstructed the way accreditation suggests or the way we image patients?

Both are informative, but the facility should evaluate the phantom under its actual clinical reconstruction, since filtered backprojection and iterative reconstructions can yield different numbers of visible rods and spheres.⁵

Key Takeaways

The Jaszczak-style phantom reports reconstructed resolution, cold-sphere contrast, and tomographic uniformity from one acquisition.¹²
The ACR-approved phantom uses cold rods of 4.8–12.7 mm and cold spheres of 9.5–31.8 mm.⁷
Collimator resolution dominates and worsens with distance: $R_{coll} = d (L_{e} + b) / L_{e}$ ; center the phantom and minimize the orbit.³⁴
Ring artifacts come from flood non-uniformity amplified by reconstruction; fix uniformity and COR before acquiring.¹³
Cold-sphere detectability follows a CNR near the Rose criterion, so adequate counts are essential.⁶
Phantom testing is quarterly and at acceptance within a physicist-directed program aligned with AAPM TG-177 and ACR accreditation.²⁷

Conclusion

The Jaszczak-style SPECT phantom endures because it tests the whole tomographic chain at once — resolution, contrast, and uniformity — and because it surfaces the reconstruction-amplified errors, especially ring artifacts and center-of-rotation offsets, that no planar test can catch. Prepared carefully, acquired under clinical settings, and analyzed quantitatively and trended over time, it is the most efficient single check a nuclear medicine physics program has for keeping SPECT images diagnostic.¹²³

How DRPS Can Help

Diagnostic Radiation Physics Services (DRPS) supports nuclear medicine and SPECT/CT facilities across Florida, Maryland, Virginia, Washington DC, California, Nevada, Pennsylvania, New York, New Jersey, and Delaware with PET/CT and nuclear medicine physics, gamma camera and SPECT acceptance testing, quarterly and annual phantom evaluation, and accreditation support for the ACR Nuclear Medicine and PET program. Our board-certified medical physicists set protocols, tolerances, and trending so that reconstructed resolution, contrast, and uniformity stay within specification.

Related Resources

References

American Association of Physicists in Medicine. Rotating Scintillation Camera SPECT Acceptance Testing and Quality Control. AAPM Report No. 22. New York: American Institute of Physics; 1987. aapm.org
Halama JR, Graham DL, Harkness BA, et al. Acceptance Testing and Annual Physics Survey Recommendations for Gamma Camera, SPECT, and SPECT/CT Systems. Report of AAPM Task Group 177. AAPM Report No. 177. College Park, MD: American Association of Physicists in Medicine; 2019. aapm.org
International Atomic Energy Agency. Quality Assurance for SPECT Systems. IAEA Human Health Series No. 6. Vienna: IAEA; 2009. iaea.org
Zanzonico P. Routine quality control of clinical nuclear medicine instrumentation: a brief review. J Nucl Med. 2008;49(7):1114-31. doi:10.2967/jnumed.107.050203. doi.org
Nichols KJ, DiFilippo FP, Palestro CJ. Texture analysis for automated evaluation of Jaszczak phantom SPECT system tests. Med Phys. 2019;46(1):262-272. doi:10.1002/mp.13289. doi.org
Tazegul TE, Polemi AM, Snyder A, Snyder C, Collins PG. Automated phantom analysis for gamma cameras and SPECT: a methodology for use in a clinical setting. J Appl Clin Med Phys. 2020;21(11):205-214. doi:10.1002/acm2.13057. doi.org
American College of Radiology. Nuclear Medicine and PET Accreditation Program. Reston, VA: ACR. acraccreditation.org