SPECT Center of Rotation: Calibration and QC

The center of rotation (COR) is the alignment between a SPECT camera's electronic image matrix and the true mechanical axis its detector heads rotate about, and keeping that alignment within tolerance is essential for artifact-free tomography. When the COR assumed during reconstruction does not match the real axis of rotation, filtered back-projection misregisters every projection: point sources reconstruct into rings, distributed activity blurs, spatial resolution and contrast fall, and on cardiac SPECT the result can look like a perfusion defect that is not there. COR is verified with an off-axis point-source acquisition and characterized using the methods in NEMA NU 1-2023 and AAPM guidance. ^{1, 2, 3}

COR is one of the highest-yield SPECT quality-control tests because the failure mode is both subtle and clinically dangerous: a modest offset does not crash the scanner or trip an obvious alarm — it quietly manufactures false findings. That is why COR calibration and verification belong in every SPECT program's routine QC. ^{3, 5}

Introduction

Tomographic reconstruction assumes it knows exactly where the rotation axis projects onto the image matrix. SPECT acquires planar projections at many angles and reconstructs them — by filtered back-projection or iterative methods — into transaxial slices. Both reconstruction families assume the projection of the mechanical axis of rotation (AOR) falls at a known, fixed column of the image matrix. The center of rotation is exactly that assumed alignment. ^{1, 4}

If the assumption is wrong — because the detector matrix, collimator, or gantry mechanics shift the effective axis — back-projected rays from opposing angles no longer intersect at the true source location. The reconstruction smears the activity around the true point. The clinical danger is that the smear is not random noise: it has structure (rings, blurring, apparent defects) that a reader can misinterpret as anatomy or pathology. ^{2, 5, 6}

This guide explains what COR is, the physics of how a COR error corrupts the reconstruction, the standardized point-source measurement, the role of gantry tilt and multi-head registration, acceptance and routine tolerances from NEMA NU 1-2023 and AAPM Report No. 177, the clinical consequences (especially in myocardial perfusion imaging), practical QC tips, and the licensing and accreditation context. DRPS performs SPECT COR and tomographic QC as part of its PET/CT and nuclear medicine physics services across Florida, Maryland, Virginia, Washington DC, California, Nevada, Pennsylvania, New York, New Jersey, and Delaware.

Topic Explanation

What is the center of rotation?

The center of rotation is the point (more precisely, the matrix column) at which the camera's electronic coordinate system coincides with the true mechanical axis of rotation of the detector heads. A perfectly aligned system has the projection of the AOR fall on the same pixel column in every view, and that column is the one the reconstruction software assumes. A COR error exists whenever the effective axis seen by the image matrix is displaced from the assumed axis used in reconstruction. ^{1, 4}

Three things can create that displacement:

• Electronic offset — the camera's image matrix is shifted relative to the mechanical axis.
• Mechanical misalignment — gantry, bearing, or detector-head mounting imperfections.
• Collimator or head effects — a collimator change or head sag that alters the effective axis along the detector.

Because all three contribute, COR is calibrated (a correction offset is stored) and then verified (the residual offset is measured and trended). ^{3, 4}

Why a small misalignment matters

SPECT reconstruction is unforgiving of registration error because it combines many views. An error that would be invisible on a single planar image becomes a structured artifact once dozens of projections are back-projected through a wrong common axis. For background on the underlying performance tests that COR sits within, see gamma camera NEMA NU-1 performance testing and the broader SPECT/CT quality control workflow. ^{1, 3}

Key Technical Principles

The point-source sinogram

A point source placed off the rotation axis at radius $r$ projects, at gantry angle $\varphi$, to a detector position that traces a sinusoid:

$$p(\varphi )=x_0+r\sin (\varphi +{\varphi }_0)$$

Here $x_0$ is the projection of the true axis — the quantity COR calibration must recover. Because the sinusoid is symmetric about $x_0$, the COR offset can be extracted from opposing projections 180° apart:

$$x_0=\frac{p(\varphi )+p(\varphi +180^{\circ })}{2}$$

In practice the system fits the full sinusoid across all angles for robustness, but the opposed-view average captures the idea: the COR is the mean position of the source's back-and-forth motion across the rotation. The measured $x_0$ is compared to the assumed reconstruction axis; the difference is the COR offset. ^{1, 4}

How a COR offset corrupts the image

If the reconstruction uses an axis displaced from the true one by ${\Delta }_{\text{COR}}$, a point source on the axis no longer back-projects to a point. For a 360° acquisition, the back-projected rays envelope a ring whose diameter is approximately:

$$D_{\text{ring}}\approx 2{\Delta }_{\text{COR}}$$

A COR offset of ${\Delta }_{\text{COR}}=9.6$ mm (for example, 2 pixels at a 4.8 mm pixel pitch) produces a ring artifact roughly:

$$D_{\text{ring}}\approx 2\times 9.6\approx 19\text{mm}$$

For distributed activity the same misregistration shows up as blurring, loss of resolution, reduced contrast, and — depending on geometry — a "tuning-fork" splitting of point-like structures. The artifact magnitude scales directly with the offset, which is why even sub-centimeter errors are clinically meaningful; published QC atlases illustrate the characteristic ring and tuning-fork appearances. ^{2, 5, 6, 9}

Gantry tilt and head sag

COR is not the only alignment that matters. A gantry tilt of angle $\theta$ displaces image data along the axis of rotation. At rotation radius $R$:

$$d=R\tan (\theta )$$

A 1° tilt at a 20 cm radius produces an axial displacement of:

$$d=200\times \tan (1^{\circ })\approx 200\times 0.0175\approx 3.5\text{mm}$$

That 3.5 mm axial shift is exactly the kind of error that appears in the y-position trend of the weekly COR point-source measurement, which is why a complete COR check examines both the in-plane offset and the axial (y) stability across the rotation and along the collimator. AAPM guidance notes that COR variation along the length of a collimator should be small — on the order of a millimeter — for a well-behaved system. ^{2, 3}

Recommended COR practice across standards

The numbers a facility applies (offset tolerance in pixels or millimeters, test frequency) come from the manufacturer's specification and the QC program built on these references; the geometry and units must match the standard and vendor exactly. ^{1, 2, 3, 4}

The measurement, step by step

A defensible COR check generally follows the same arc:

1. Place a point source (commonly Co-57 or a small Tc-99m source) off the axis of rotation at the manufacturer-specified position.
2. Acquire a full SPECT orbit per the vendor/NEMA protocol for each detector head.
3. Fit the source position versus angle to recover $x_0$ (in-plane) and track the y-position (axial).
4. Compare the measured offset to tolerance; apply or update the stored calibration if required.
5. For multi-head systems, verify head-to-head registration so all detectors share one effective axis.
6. Document results and trend them over time to catch drift before it becomes clinical.

Clinical Impact

The defining clinical hazard of COR error is the false-positive study. Because the artifact has structure, it does not merely add noise — it can imitate disease.

• Myocardial perfusion SPECT. COR error is a recognized source of myocardial perfusion artifacts. A misalignment can blur the ventricular wall or create apparent reduced-count regions that mimic ischemia or infarct, raising the false-positive rate of a high-volume test. DePuey's classic review lists center-of-rotation error among the instrumentation artifacts that must be controlled or detected to keep myocardial SPECT reliable, and Baron and Chouraqui place COR within the myocardial SPECT quality-assurance program. ^{5, 6} For modality-specific QC, see cardiac SPECT MPI quality control.
• Bone, brain, and tumor SPECT. Loss of resolution and contrast from a COR offset reduces lesion detectability and can distort the apparent size and location of focal uptake.
• Quantitative SPECT and theranostics. Modern quantitative SPECT/CT — including post-therapy Lu-177 dosimetry — depends on accurate spatial registration and alignment with CT for attenuation correction. A COR error degrades both localization and quantitative recovery, undermining the numbers used for dosimetry. This connects directly to Lu-177 theranostics dosimetry and quantitative workflows.
• Multi-head registration. On dual- and triple-head systems, mutual head misregistration can produce artifacts even when each head's individual COR appears acceptable, so head-to-head alignment is part of the clinical risk picture. ^{1, 3}

Practical Optimization Tips

1. Respect the vendor geometry

COR tolerances are expressed in pixels or millimeters for a specific matrix, zoom, and source position. Apply the limit in the same units and geometry the manufacturer specifies; a "0.5 pixel" tolerance means different millimeter values at different zooms. ^{1, 3}

2. Trend, do not just pass/fail

A single in-tolerance COR result is less informative than a trend. Plotting the offset and the axial y-position over weeks reveals slow mechanical drift, bearing wear, or head sag before it crosses the action limit. Treat COR like any constancy test: watch the slope, not just the latest point. ³

3. Re-check after anything mechanical

Always repeat COR after a collimator change, detector service, gantry collision, or software/calibration update. These are precisely the events that move the effective axis, and a stale calibration is worse than none because it is trusted. ^{3, 4}

4. Use an appropriate source and counts

Follow the vendor and NEMA recommendations for source type, activity, and counts. Too few counts make the sinusoid fit noisy; an incorrectly positioned source biases the offset. Co-57 point sources are convenient because of their longevity and photon energy near Tc-99m. ^{1, 4}

5. Verify multi-head registration explicitly

Do not assume that good single-head COR guarantees good head-to-head alignment. Check registration across all detectors so the reconstruction sees one consistent axis. ^{1, 3}

Common pitfalls to avoid

• Trusting an auto-calibration without verification. Automated COR correction still needs an independent check and trend.
• Applying a pixel tolerance at the wrong zoom. The millimeter meaning of a pixel changes with matrix and magnification.
• Skipping COR after collimator swaps. The effective axis can move with the collimator.
• Ignoring the axial (y) trend. Gantry tilt and head sag show up axially, not just in-plane.
• Reading subtle cardiac defects without QC context. An unrecognized COR artifact can be over-called as ischemia.

Regulatory Considerations

COR calibration is part of the quality-assurance framework expected of a licensed nuclear medicine program, even though no single regulation prints a COR tolerance. The obligation flows from the requirement to operate instrumentation properly and to maintain a radiation safety and QA program.

• 10 CFR Part 35 — Medical Use of Byproduct Material. Governs the medical use program, authorized users, and the radiation safety framework within which imaging-instrument QA, including SPECT performance testing, is maintained. Agreement States administer equivalent rules. ⁷
• 10 CFR Part 20 — Standards for Protection Against Radiation. Sets the broader radiation protection and ALARA context; reliable imaging QC supports keeping repeat studies and unnecessary dose to a minimum. ⁸
• NEMA NU 1-2023 and AAPM Reports 22 and 177. The technical backbone for how COR is measured and what acceptance and routine testing should include. Accreditation programs and manufacturers reference these for performance expectations. ^{1, 2, 3}
• IAEA Human Health Series No. 6. Internationally used, detailed QA procedures for SPECT, including center of rotation and alignment of axes. ⁴

Of the states DRPS serves, Florida, Maryland, Virginia, California, Nevada, Pennsylvania, New York, and New Jersey are NRC Agreement States that license medical use under their own radiation-control programs, while Washington, DC and Delaware are regulated directly by the NRC for byproduct material. Accreditation bodies (for example, for nuclear medicine and cardiac imaging) and the facility's license conditions determine the specific QC documentation expected; confirm requirements with the authority having jurisdiction. For preparation guidance, see preparing for an NRC inspection. ^{7, 8}

Frequently Asked Questions (FAQs)

What is the center of rotation in SPECT?

The center of rotation (COR) is the alignment between the gamma camera's electronic image matrix and the true mechanical axis about which the detector heads rotate. When the assumed center in the projection data does not match the real rotation axis, back-projection misregisters the projections and the reconstruction is blurred or artifacted.

What artifacts does a center-of-rotation error cause?

A small COR error blurs the reconstruction and degrades spatial resolution and contrast. A larger error produces a characteristic ring or doughnut artifact from a point source — sometimes called a tuning-fork artifact — and on cardiac SPECT it can create apparent perfusion defects that mimic disease and cause false-positive reads.

How is the center of rotation measured?

A point or line source is imaged off-axis through a full SPECT rotation. The source's position in each projection traces a sinusoid; the COR offset is found from the mean of opposing projections or by fitting the sinusoid. NEMA NU 1-2023 defines the standardized method, and most modern systems include an automated COR calibration and check.

What is an acceptable COR tolerance?

Tolerances are set by the manufacturer and the facility's QC program and are commonly on the order of a fraction of a pixel; many programs investigate offsets approaching about half a pixel. Because pixel size varies with matrix and zoom, the limit must be applied in the units (pixels or millimeters) and geometry specified by the standard and the vendor.

How often should center of rotation be checked?

COR is verified at acceptance testing and then on a routine schedule — commonly weekly to monthly depending on the system and the QC program — and always after detector or collimator service, a crash or collision, or a software update. AAPM Report No. 177 provides acceptance-testing and routine-survey recommendations for gamma camera, SPECT, and SPECT/CT systems.

Does center-of-rotation error affect multi-head cameras differently?

Yes. Multi-detector systems must also be registered to each other. If two or three heads do not share the same effective rotation axis, their projections are mutually misaligned, which degrades resolution and can introduce artifacts even when each head's individual COR looks acceptable. Head-to-head registration is checked alongside COR.

Is center of rotation important for quantitative SPECT?

Very. Quantitative SPECT and SPECT/CT depend on accurate spatial registration so that activity is reconstructed at the correct location and aligned with the CT for attenuation correction. A COR error degrades both spatial accuracy and the quantitative recovery that modern theranostic dosimetry and quantitative imaging rely on.

Key Takeaways

• COR is an alignment, not a place. It is the match between the camera's image matrix and the true mechanical axis of rotation used by the reconstruction.
• Small offsets cause structured artifacts. A point source becomes a ring of diameter roughly $2{\Delta }_{\text{COR}}$, and distributed activity blurs and loses contrast.
• The point-source sinusoid is the test. The COR offset is the mean position of an off-axis source across the rotation, recovered from opposed views or a full fit.
• Gantry tilt and head sag matter too. A 1° tilt at 20 cm shifts data ~3.5 mm axially; a full check looks in-plane and axially, and across the collimator.
• The clinical danger is false positives. COR error is a known source of myocardial perfusion artifacts and degrades quantitative SPECT.
• Trend and re-check. Watch the COR trend and always repeat after collimator changes, service, collisions, or software updates.

Conclusion

Center-of-rotation calibration is a small test with outsized clinical stakes. Because SPECT combines many projections through one assumed axis, a misalignment that would be invisible on a planar image becomes a structured artifact in the reconstruction — and that structure can imitate disease. By understanding the point-source sinogram, applying the NEMA NU 1-2023 and AAPM methods, respecting vendor geometry, trending results, and re-checking after any mechanical or software change, a nuclear medicine program can keep COR within tolerance and keep its tomographic images trustworthy. As quantitative SPECT/CT and theranostic dosimetry raise the bar on spatial accuracy, disciplined COR QC only becomes more important.

How DRPS Can Help

Diagnostic Radiation Physics Services supports nuclear medicine programs with SPECT and SPECT/CT acceptance testing, center-of-rotation and tomographic-uniformity evaluation, multi-head registration checks, annual physics surveys, and QC-program design, delivered by board-certified medical physicists as part of PET/CT and nuclear medicine physics and accreditation support.

DRPS serves facilities across our service locations, including Florida, Maryland, Virginia, Washington DC, California, Nevada, New York, Pennsylvania, New Jersey, and Delaware, and integrates SPECT QC with broader medical physicist consulting.

Reliable tomography starts with a correct axis. Keeping COR in tolerance is one of the most direct ways to protect both image quality and diagnostic accuracy.

Related Resources

• Gamma camera NEMA NU-1 performance testing
• SPECT/CT quality control
• Cardiac SPECT MPI quality control
• Gamma camera collimator selection
• Lu-177 theranostics dosimetry
• PET/CT and nuclear medicine physics
• Accreditation support

References

1. National Electrical Manufacturers Association. NEMA Standards Publication NU 1-2023 — Performance Measurements of Gamma Cameras. Rosslyn, VA: NEMA; 2023. nema.org
2. American Association of Physicists in Medicine. Rotating Scintillation Camera SPECT Acceptance Testing and Quality Control. AAPM Report No. 22. College Park, MD: AAPM; 1987. aapm.org
3. Halama JR, Madsen MT, Dickson J, et al. Acceptance Testing and Annual Physics Survey Recommendations for Gamma Camera, SPECT, and SPECT/CT Systems. AAPM Report No. 177. College Park, MD: AAPM; 2019. aapm.org
4. International Atomic Energy Agency. Quality Assurance for SPECT Systems. IAEA Human Health Series No. 6. Vienna: IAEA; 2009. iaea.org
5. DePuey EG. How to detect and avoid myocardial perfusion SPECT artifacts. J Nucl Med. 1994;35(4):699-702. PubMed
6. Baron JM, Chouraqui P. Myocardial single-photon emission computed tomographic quality assurance. J Nucl Cardiol. 1996;3(2):157-166. doi:10.1016/s1071-3581(96)90008-2. PubMed
7. U.S. Nuclear Regulatory Commission. 10 CFR Part 35 — Medical Use of Byproduct Material. ecfr.gov
8. U.S. Nuclear Regulatory Commission. 10 CFR Part 20 — Standards for Protection Against Radiation. ecfr.gov
9. International Atomic Energy Agency. Quality Control Atlas for Scintillation Camera Systems. Vienna: IAEA; 2003. iaea.org