What is multiple-window spatial registration (MWSR)?

Multiple-window spatial registration is a gamma camera performance parameter that measures how far the image of a point source shifts when it is acquired in energy windows centered on different photopeaks. It is quantified as the maximum displacement of the imaged centroid between windows and is defined in the NEMA NU 1 standard. Good MWSR means a source appears in the same location regardless of which photopeak window records it.

Why does the imaged position depend on photon energy?

A gamma camera estimates each event's position from the weighted signals of its photomultiplier tubes. Because the light distribution and tube response change with photon energy, the position estimate can shift slightly between a low-energy and a high-energy photopeak if the camera is not well tuned. MWSR measures that residual energy-dependent shift.

Which radionuclides make MWSR important?

MWSR matters most for radionuclides imaged with more than one photopeak window, such as gallium-67 with photopeaks near 93, 184, and 300 keV and indium-111 with photopeaks near 171 and 245 keV, and for dual-isotope studies that combine two radionuclides such as technetium-99m with iodine-123 or thallium-201. In all of these, poor registration between windows blurs the combined image.

What is an acceptable MWSR value?

MWSR is reported as a displacement in millimeters, and acceptance values are commonly specified on the order of about 2 mm or less within the central field of view, with the exact limit set by the manufacturer specification and the NEMA NU 1 measurement conditions. The physicist compares the measured value against the manufacturer specification at acceptance and tracks it over time.

How is MWSR measured?

A point source of a multi-energy radionuclide, typically gallium-67, is imaged at defined locations in the field of view. The centroid of the counts is computed separately in the energy window for each photopeak, and MWSR is the maximum distance between those centroids. The test is part of NEMA NU 1 acceptance testing and is repeated when the camera is serviced or its energy calibration changes.

Does MWSR relate to dual-isotope crosstalk?

They are related but distinct. MWSR is a geometric registration check between energy windows, while crosstalk and downscatter describe photons from one radionuclide being detected in the other's window. Both degrade dual-isotope images, and both are managed through careful energy-window setup, scatter correction such as the triple-energy-window method, and routine quality control.

Who should perform gamma camera MWSR testing?

MWSR is part of acceptance testing and periodic performance evaluation performed by a qualified or board-certified medical physicist, following NEMA NU 1 methodology and the facility's accreditation requirements. The technologist's daily quality control supports it through energy peaking and uniformity checks, but the formal MWSR measurement is a physics task.

Gamma Camera Multiple-Window Spatial Registration

Multiple-window spatial registration (MWSR) checks that a gamma camera places a source at the same location no matter which energy window records it. Because a scintillation camera estimates each photon's position from the energy-dependent signals of its photomultiplier tubes, the imaged position can shift slightly between a low-energy and a high-energy photopeak — and MWSR measures that shift. It is a NEMA performance parameter that becomes clinically important for multi-energy radionuclides such as gallium-67 and indium-111 and for dual-isotope studies, where poor registration blurs the combined image and corrupts quantification.¹²

MWSR is one of those quality-control parameters that is invisible on a single-photopeak technetium-99m scan but decisive on a gallium-67 or octreotide study. This guide explains why imaged position depends on energy, how the NEMA test is performed, what tolerances apply, how MWSR relates to dual-isotope crosstalk, and the quality-control routine that keeps a multi-energy imaging program defensible.¹²

Introduction

A gamma camera is, at heart, a position-estimating machine, and that estimate is energy dependent. When a photon interacts in the sodium iodide crystal, the resulting light spreads across an array of photomultiplier tubes, and the camera computes the event position from the weighted tube signals. The width and shape of that light distribution change with photon energy, so a camera that is perfectly registered at 140 keV can place a 93 keV or a 300 keV event a fraction of a millimeter away.¹³

For single-isotope, single-window imaging — the technetium-99m studies that make up most of a nuclear-medicine schedule — this does not matter, because only one window is in use. It matters a great deal when two or three windows are combined into one image, either for a single multi-energy radionuclide or for two radionuclides acquired at once. MWSR is the parameter that quantifies whether those windows agree on where a source is.¹²

This guide covers the detector physics behind energy-dependent positioning, the NEMA NU 1 MWSR test and its tolerances, a worked example of energy-window and registration arithmetic, the clinical consequences of misregistration and crosstalk, and the practical quality control that a program imaging gallium-67, indium-111, or dual isotopes should have in place.

Topic Explanation

How a gamma camera assigns position

A conventional scintillation (Anger) camera uses a single large sodium iodide thallium-activated crystal, NaI(Tl), viewed by an array of photomultiplier tubes. Each detected photon produces a burst of scintillation light, and the camera estimates the event position from the relative tube signals and the event energy from their sum. The position and energy signals are linked, and the electronics apply energy-, linearity-, and uniformity-correction maps so that the final image is spatially accurate and flat.¹³

Those correction maps are tuned, and they depend on energy. If the energy calibration drifts or the correction for a given photopeak is imperfect, the estimated position of an event can depend slightly on which photopeak window it falls in. The visible result is that a point source imaged in two different windows appears in two slightly different places — a registration error between windows.¹³

Energy windows and photopeaks

A gamma camera records an event only if its measured energy falls within a selected energy window, usually expressed as a percentage of the photopeak energy centered on that peak. A radionuclide with a single dominant gamma emission, such as technetium-99m at 140 keV, uses one window. Several clinically important radionuclides emit at more than one energy and are imaged with multiple windows summed together:

Gallium-67 — photopeaks near 93, 184, and 300 keV (with a further peak near 388 keV), commonly imaged with two or three windows.⁶⁷
Indium-111 — photopeaks near 171 and 245 keV, imaged with both windows.
Thallium-201 — mercury X-rays near 69-80 keV plus a gamma near 167 keV.

Dual-isotope studies go a step further and combine two radionuclides acquired simultaneously — for example technetium-99m with iodine-123 (159 keV), or technetium-99m with thallium-201 — so that two or more windows are open at once and must register to one another.²⁸

Why registration between windows matters

When windows are summed, any displacement between them smears the combined image. A point that is sharp in each individual window becomes a short streak or a broadened blob once the windows are added, degrading resolution and contrast. In quantitative and dual-isotope work, misregistration also mixes signal between the two radionuclide images. MWSR is the measurement that tells a physicist whether the camera's windows agree on position closely enough that summing them is safe.¹²

Key Technical Principles

Energy-window width and placement

An energy window of fractional width $f$ centered on a photopeak energy $E_{0}$ spans:

Δ E = f \cdot E_{0}, [E_{0} - \frac{Δ E}{2}, E_{0} + \frac{Δ E}{2}]

For a symmetric 20% window at the technetium-99m photopeak of 140.5 keV:

Δ E = 0.20 \times 140.5 = 28.1 keV, window \approx 126 - 155 keV

Window width is a compromise: a wider window captures more true photopeak counts (higher sensitivity) but admits more scattered photons (lower contrast), while a narrower window does the reverse. For multi-energy imaging the choice interacts with registration and crosstalk, which is why the optimal width is not the same for every photopeak.¹⁶

Energy resolution

The ability of the camera to distinguish photopeaks is set by its energy resolution, defined as the full width at half maximum (FWHM) of a photopeak divided by its energy:

R (%) = \frac{FWHM}{E _{0}} \times 100

A modern NaI(Tl) gamma camera typically achieves an energy resolution of roughly 9-10% at 140 keV. For example, a measured photopeak FWHM of 13.7 keV at 140.5 keV gives:

R = \frac{13.7}{140.5} \times 100 \approx 9.8%

Good energy resolution lets the system separate closely spaced photopeaks — important when two radionuclides emit near one another — and is a prerequisite for stable registration between windows.³

The MWSR test

NEMA NU 1 defines MWSR as the maximum displacement of the imaged centroid of a point source measured across the energy windows used for a multi-energy radionuclide.¹ In practice:

A point source of a multi-energy radionuclide — gallium-67 is the standard choice because its photopeaks span a wide energy range — is positioned at defined points in the field of view.
The system acquires the source simultaneously in a separate energy window centered on each photopeak.
The centroid (center of counts) of the source image is computed in each window.
MWSR is the largest distance between any pair of those centroids.

For centroids at positions $(x_{i}, y_{i})$ in two windows, the displacement is:

d = (x_{1} - x_{2})^{2} + (y_{1} - y_{2})^{2}

As a worked example, suppose the gallium-67 source centroid is measured at the 93 keV window and the 300 keV window and the centroid coordinates differ by $Δ x = 0.9$ mm and $Δ y = 1.2$ mm. The registration error between those windows is:

d = 0. 9^{2} + 1. 2^{2} = 0.81 + 1.44 = 2.25 = 1.5 mm

A value of 1.5 mm would pass a typical acceptance specification of about 2 mm or less in the central field of view, though the exact limit is set by the manufacturer and the NEMA measurement conditions.¹ Registration tends to degrade toward the edge of the field of view, so the measurement is made at multiple locations, not just the center.

Multi-energy radionuclides and their windows

Radionuclide	Photopeaks imaged (keV)	Typical windowing	Why MWSR matters
Technetium-99m	140	Single 15-20% window	Single window; MWSR not in play
Gallium-67	93, 184, 300 (388)	Two or three windows summed	Wide energy span stresses registration
Indium-111	171, 245	Two windows summed	Octreotide/white-cell imaging rely on both
Thallium-201	69-80 (X-rays), 167	Low window plus gamma	Low-energy window sensitive to tuning
Tc-99m + I-123	140 and 159	Dual isotope, two windows	Close peaks; registration and crosstalk

The pattern is clear: the more windows that must be summed, and the wider the energy range they span, the more a small energy-dependent position shift is amplified in the final image.²⁶⁸

Clinical Impact

MWSR is a quiet parameter until a multi-energy study goes soft. On a gallium-67 scan for infection or lymphoma, or an indium-111 octreotide or white-blood-cell study, misregistration between photopeak windows blurs lesions, lowers contrast, and can make a small focus harder to detect against background. Because these studies already have modest count statistics, any avoidable loss of resolution matters.²⁶

In dual-isotope imaging — simultaneous technetium-99m and iodine-123 for parathyroid localization, or technetium-99m and thallium-201 protocols — two separate problems compound. Poor MWSR shifts one radionuclide's image relative to the other, and energy-window crosstalk lets photons from the higher-energy radionuclide scatter down into the lower window. A gallium-67 study investigated with two versus three photopeak windows, for instance, shows measurable differences in uniformity, spatial resolution, and lesion contrast depending on how the windows are chosen and weighted.⁶ For quantitative dual-isotope SPECT, scatter and crosstalk corrections such as the triple-energy-window method are needed to recover accurate activity, and they assume the underlying windows are well registered.⁸

The practical message for a nuclear-medicine program is that a camera can pass every daily technetium-99m check and still be unfit for gallium-67 or dual-isotope work if its multi-window registration has drifted. MWSR is the test that catches that gap.¹²

Practical Optimization Tips

Peak and calibrate before multi-energy work

Energy-dependent positioning errors grow when the energy calibration is off, so daily energy peaking and periodic full energy calibration are the foundation of good registration. Before a gallium-67 or dual-isotope study, confirm that each photopeak window is correctly centered on its peak rather than relying on a single technetium-99m peaking.³

Choose window widths deliberately

Do not simply apply a default 20% window to every photopeak. For gallium-67, published work shows that optimizing window widths per photopeak — for example a tighter window on the low-energy peak and appropriately set windows on the higher peaks — improves contrast and lesion detectability compared with naive settings.⁶ Match the windowing to the task and document it in the protocol.

Select the right collimator

Multi-energy radionuclides with high-energy photons, such as gallium-67 and indium-111, need medium- or high-energy collimators to control septal penetration. Imaging these radionuclides through a low-energy collimator admits penetrating photons that degrade resolution and effectively worsen apparent registration, so collimator selection is part of getting clean multi-window images.⁷

Test MWSR at acceptance and after service

Measure MWSR at acceptance against the manufacturer specification, and repeat it after any event that can perturb energy calibration or correction maps — a detector service, a photomultiplier-tube or electronics replacement, or a correction-map update. Track the values so that a slow drift is visible before it reaches a study. Support the formal MWSR measurement with routine uniformity checks performed at the relevant energies, not only at 140 keV.¹³

Apply scatter and crosstalk corrections for dual isotopes

When two radionuclides are acquired together, plan for crosstalk from the start: use energy windows with adequate separation, apply a validated scatter-correction method such as the triple-energy-window approach, and verify quantitative accuracy on a phantom before trusting patient numbers.⁸

Regulatory Considerations

Gamma camera performance QC is driven by accreditation and manufacturer standards rather than by a specific federal test list. The imaging itself uses byproduct material, so it falls under NRC or Agreement State medical-use rules, but those rules focus on radiation safety and the medical-use program rather than prescribing camera performance tests.

Radioactive material program. Possession and medical use of gallium-67, indium-111, and other imaging radionuclides fall under 10 CFR Part 35 (or the equivalent Agreement State program), with dose limits under 10 CFR Part 20 and licensing guidance in NRC NUREG-1556 Volume 9.⁹¹⁰
Performance standards. The methodology and definitions for MWSR and the other gamma camera performance parameters come from the NEMA NU 1 standard, with IAEA guidance providing practical quality-control procedures.¹²⁴
Accreditation. Accreditation programs and the facility's own quality-management plan are usually what require documented acceptance testing and periodic performance evaluation, including the multi-energy parameters relevant to gallium-67 and indium-111 imaging.
Who performs the work. Acceptance testing and performance evaluation, including MWSR, are the responsibility of a qualified or board-certified medical physicist. DRPS supports facilities across Florida, Maryland, Virginia, Washington DC, California, Nevada, Pennsylvania, New York, New Jersey, and Delaware; confirm specific requirements with the authority having jurisdiction and the facility's accrediting body.

Frequently Asked Questions (FAQs)

Is MWSR tested with technetium-99m?

No. MWSR requires a radionuclide with multiple photopeaks spanning a range of energies, so gallium-67 is the standard test source because its 93, 184, and 300 keV peaks exercise the full energy range. A single-peak radionuclide such as technetium-99m cannot reveal a registration error between windows.¹⁶

How often should MWSR be measured?

MWSR is measured at acceptance and repeated whenever the energy calibration or correction maps may have changed — after detector service, electronics or photomultiplier-tube work, or a correction-map update — and at the periodic performance-evaluation interval set by the physicist and the accrediting body.¹³

What is the difference between MWSR and uniformity?

Uniformity describes how evenly the camera responds across the field of view at a single energy, while MWSR describes how consistently it places a source across different energy windows. Both are NEMA performance parameters, and both should be checked at the energies a program actually images.¹

Can poor MWSR be corrected in software?

The primary fix is correct energy calibration and correction-map tuning on the camera, which is a service and physics task. Image-processing corrections cannot reliably undo a registration error that has already smeared summed-window data, which is why the measurement and the tuning come first.¹³

Why do gallium-67 and indium-111 need special attention?

Both are imaged with multiple photopeak windows and emit high-energy photons, so they stress both registration (MWSR) and collimation (septal penetration). Getting clean images requires correct window setup, a medium- or high-energy collimator, and verified MWSR.⁶⁷

Key Takeaways

MWSR measures the maximum displacement of a point source's imaged centroid across energy windows, a NEMA NU 1 performance parameter.¹
The imaged position is energy dependent because the camera estimates position from energy-dependent photomultiplier-tube signals.³
MWSR matters for multi-energy radionuclides such as gallium-67 (93, 184, 300 keV) and indium-111 (171, 245 keV) and for dual-isotope studies.²⁶
Acceptance values are commonly about 2 mm or less in the central field of view, set by the manufacturer specification and NEMA conditions.¹
Window-width choice, energy calibration, collimator selection, and crosstalk correction all interact with registration in multi-window imaging.⁶⁷⁸
A camera can pass daily technetium-99m checks and still be unfit for multi-energy work if MWSR has drifted — which is why the physics test exists.¹²

Conclusion

Multiple-window spatial registration is a small number with an outsized effect on the studies that depend on it. For the routine technetium-99m workload it is dormant, but for gallium-67, indium-111, and dual-isotope imaging it determines whether summed-window images stay sharp and whether quantification can be trusted. Because the error is rooted in energy-dependent position estimation, the remedy is physics-driven: correct energy calibration, deliberate window setup, appropriate collimation, and a measured, tracked MWSR against the manufacturer specification. A program that treats MWSR as part of its acceptance and performance-evaluation routine is a program whose multi-energy images can be believed.¹²³

How DRPS Can Help

Diagnostic Radiation Physics Services (DRPS) supports nuclear-medicine and PET/CT facilities across Florida, Maryland, Virginia, Washington DC, California, Nevada, Pennsylvania, New York, New Jersey, and Delaware with PET/CT and nuclear medicine physics, accreditation support, and medical physicist consulting — including NEMA NU 1 acceptance testing, gamma camera performance evaluation, multi-energy window optimization for gallium-67 and indium-111, dual-isotope crosstalk assessment, and documentation prepared by board-certified medical physicists.

Multi-energy imaging quality is won at acceptance and maintained through routine quality control. DRPS helps facilities build and document the testing that keeps gallium-67, octreotide, and dual-isotope studies sharp and quantitatively reliable.

Related Resources

References

National Electrical Manufacturers Association. Performance Measurements of Gamma Cameras. NEMA Standards Publication NU 1-2018. Rosslyn, VA: NEMA; 2018. nema.org
Zanzonico P. Routine quality control of clinical nuclear medicine instrumentation: a brief review. Journal of Nuclear Medicine. 2008;49(7):1114-1131. doi:10.2967/jnumed.107.050203. doi.org
International Atomic Energy Agency. Quality Assurance for SPECT Systems. IAEA Human Health Series No. 6. Vienna: IAEA; 2009. iaea.org
International Atomic Energy Agency. Quality Control Atlas for Scintillation Camera Systems. Vienna: IAEA; 2003. iaea.org
American Association of Physicists in Medicine. Rotating Scintillation Camera SPECT Acceptance Testing and Quality Control. AAPM Report No. 22. New York, NY: AAPM; 1987. aapm.org
Ali L, Khalil M, Hadi N. Revisiting gallium-67 imaging: investigation of the energy photopeaks. Nuclear Medicine Communications. 2010;31(12):1068-1074. doi:10.1097/MNM.0b013e3283400255. doi.org
Moore SC, Foley Kijewski M, El Fakhri G. Collimator optimization for detection and quantitation tasks: application to gallium-67 imaging. IEEE Transactions on Medical Imaging. 2005;24(10):1347-1356. doi:10.1109/TMI.2005.857211. doi.org
Esquinas PL, Rodriguez-Rodriguez C, Esposito TVF, et al. Dual SPECT imaging of 111In and 67Ga to simultaneously determine in vivo the pharmacokinetics of different radiopharmaceuticals: a quantitative tool in pre-clinical research. Physics in Medicine and Biology. 2018;63(23):235029. doi:10.1088/1361-6560/aaef63. doi.org
U.S. Nuclear Regulatory Commission. 10 CFR Part 35, Medical Use of Byproduct Material. nrc.gov
U.S. Nuclear Regulatory Commission. Consolidated Guidance About Materials Licenses: Program-Specific Guidance About Medical Use Licenses. NUREG-1556, Volume 9, Revision 3. nrc.gov