Skip to main content

Gamma Camera Energy Windows & Scatter Correction

By Troy Zhou, PhD, DABR, DABSNM
• 16 min read

The energy window is one of the most consequential settings on a gamma camera, and one of the easiest to get wrong. Where the photopeak window sits, how wide it is, and whether scatter subwindows are added together decide how much scattered radiation contaminates the image, how much image contrast survives, and whether a quantitative SPECT value can be defended.

Every planar scintigraphy and SPECT study begins with a decision about which detected photons to keep. A gamma camera measures the energy of each interaction in its detector, and the energy window is the acceptance gate: photons whose measured energy falls inside the window are binned into the image, and the rest are discarded. Because Compton scatter lowers a photon's energy, the window is the camera's primary — and sometimes only — line of defense against scatter before reconstruction 1, 4.

This guide explains how photopeak windows are set, why width is a trade-off rather than a target, how multi-photopeak radionuclides are handled, and how triple-energy-window scatter correction turns extra subwindows into a quantitative correction. DRPS supports nuclear medicine programs through its PET/CT and nuclear medicine physics services across Florida, Maryland, Virginia, Washington DC, California, Nevada, Pennsylvania, New York, New Jersey, and Delaware.

Introduction

A gamma camera does not image photons directly; it images the pulses those photons create in a scintillation detector. When a gamma ray deposits energy in the sodium iodide crystal, the resulting scintillation light is converted by photomultiplier tubes into an electrical pulse whose height is proportional to the deposited energy. The pulse-height analyzer then asks a single question of every event: is this pulse inside the energy window?

The answer determines the image. Set the window too low and the camera accepts a flood of Compton-scattered photons that blur contrast. Set it off-center and the camera throws away good photopeak counts, driving up noise and distorting uniformity. Add well-placed scatter subwindows and the same camera can estimate and subtract the scatter it does accept, recovering contrast and supporting quantification 1, 4, 8.

Energy windowing is therefore not a one-time factory setting. It is a daily quality-control checkpoint, a per-radionuclide acquisition parameter, and the physical foundation of every scatter-correction method a department uses. For the companion measurement of how sharply the camera can separate those energies, see our guide to gamma camera energy resolution QC.

Topic Explanation

What the energy window actually selects

A radionuclide emits gamma rays at discrete, characteristic energies. Tc-99m, the workhorse of nuclear medicine, emits a 140.5 keV gamma ray 10. In an ideal detector, every unscattered 140.5 keV photon would register at exactly 140.5 keV. Real detectors blur that line into a roughly bell-shaped photopeak because of statistical variation in light production and collection. The width of that blur is the camera's energy resolution.

The energy window is a band placed over the photopeak. A symmetric window centered on the photopeak and expressed as a percentage of the photopeak energy is the standard approach. A 20 percent window on Tc-99m spans:

Photons that underwent Compton scatter in the patient arrive with reduced energy. A 140 keV photon scattered through 50 degrees loses roughly 10 keV; one scattered through larger angles loses more. Much of that scattered radiation falls below the lower window threshold and is rejected — but not all of it. Small-angle scatter loses little energy and leaks into the photopeak window, which is exactly the problem scatter correction is designed to address 4.

Why energy resolution sets the floor

How tightly a window can reject scatter depends on how sharp the photopeak is. Energy resolution is quantified as the full width at half maximum (FWHM) of the photopeak, expressed as a percentage of the photopeak energy:

A conventional sodium iodide gamma camera has an energy resolution near 9 to 10 percent at 140 keV 4. That means the photopeak FWHM is about 14 keV wide. A 20 percent (28 keV) window is therefore roughly twice the FWHM — wide enough to capture the great majority of unscattered photopeak counts without chasing the statistical tails. Cameras built on cadmium-zinc-telluride (CZT) solid-state detectors achieve substantially better energy resolution, which allows narrower windows that reject more scatter for the same photopeak yield 7. The relationship between resolution and window width is the single most important reason a window appropriate for one camera technology is not automatically appropriate for another.

For how the photopeak and its width are measured and tracked as a routine test, see gamma camera energy resolution QC and the broader instrument QC program in gamma camera uniformity QC.

Key Technical Principles

Window width is a sensitivity–scatter trade-off

Widening the window increases the number of recorded counts, which lowers Poisson noise and brightens the image. It also admits more scattered photons, which lowers contrast and biases any quantitative measurement upward. Narrowing the window does the reverse. There is no setting that optimizes both; the window width is chosen to balance them, and international guidance for Tc-99m converges on photopeak windows in the 15 to 20 percent range 8.

The table below summarizes representative photopeak and window settings for common single-photon radionuclides. The photopeak energies are nuclear constants 10; the window settings are representative starting points that each site should confirm against its camera, its acquisition software, and the relevant procedure standard.

Radionuclide Primary photopeak(s) (keV) Representative photopeak window Windowing note
Tc-99m 140.5 15–20% symmetric (≈126–154 keV at 20%) Single-peak; the reference case for most QC
I-131 364 15–20% symmetric High energy; collimator septal penetration matters more than low-energy scatter
Tl-201 69–83 (Hg X-rays), 167 ~30% on the 69–83 keV cluster, added window on 167 keV Low-energy X-rays dominate counts; downscatter is significant
Ga-67 93, 185, 300 15–20% on each of the three peaks Multi-window acquisition; higher peaks contaminate lower-peak windows
In-111 171, 245 15–20% on each peak Dual-window acquisition

The entries are representative, not a substitute for the window settings validated on a specific camera for a specific procedure. Low-energy radionuclides such as Tl-201 carry a large downscatter and septal-penetration component, so their windowing and correction strategy differs from a clean single-peak emitter like Tc-99m 4, 6.

Asymmetric and multiple windows

A symmetric window is not always optimal. Because the scatter contribution rises toward lower energies, some protocols use an asymmetric window shifted slightly toward higher energies to reject more scatter at the cost of some photopeak counts. Radionuclides with several photopeaks require multiple simultaneous windows, and when more than one radionuclide is imaged at once — a dual-isotope study — photons from the higher-energy emitter scatter down into the lower-energy window, producing crosstalk that must be corrected separately 6. See dual-isotope SPECT imaging for how that crosstalk is managed, and gamma camera multiple-window spatial registration QC for the QC test that confirms images from different windows line up.

Triple-energy-window scatter correction

The most widely implemented energy-based scatter correction is the triple-energy-window (TEW) method of Ogawa and colleagues 1. It adds two narrow subwindows immediately adjacent to the photopeak window — one just below, one just above. The counts in those subwindows sample the scatter spectrum on each side of the photopeak, and a trapezoidal estimate of the scatter under the photopeak window is computed pixel by pixel:

Here and are the counts in the lower and upper subwindows, and are the subwindow widths, and is the photopeak window width. The scatter-corrected photopeak counts are then .

Worked example. Consider a Tc-99m acquisition with a 28 keV photopeak window (126–154 keV) and two 7 keV subwindows. In one pixel the camera records counts, counts, and counts. The estimated scatter in the photopeak window is:

so the primary (unscattered) estimate is counts, implying a scatter fraction of about 35 percent in that pixel. TEW is attractive because it needs no extra acquisition time and adapts to the actual object, but the narrow subwindows are themselves noisy, so the estimate is often smoothed before subtraction 1, 4. More elaborate model-based and Monte-Carlo scatter corrections can be more accurate but are more complex and slower 4. For a department's practical implementation, see SPECT scatter correction.

Clinical Impact

Contrast, lesion detection, and uniformity

Scatter that leaks into the photopeak window adds a smooth, diffuse background to the image. That background reduces the contrast between a lesion and its surroundings, which is precisely the signal a reader is trying to detect. Correctly centered windows and effective scatter rejection improve lesion contrast; a window that has drifted off the photopeak, or is set too wide, degrades it. A mis-centered window also degrades field uniformity, because the fraction of the photopeak captured varies across the detector face 4.

Quantitative SPECT

Modern SPECT/CT can report an absolute activity concentration and standardized uptake value, but only if the counts feeding the reconstruction are clean. Scattered photons counted in the photopeak window inflate the apparent activity and bias quantification. With CT-based attenuation correction and a validated scatter correction in place, Tc-99m quantitative SPECT can achieve accuracy within about plus or minus 10 percent of the true concentration 3. That accuracy collapses if the energy window is wrong or the scatter correction is absent. Energy windowing is, in this sense, the first link in the quantitative chain described in quantitative SPECT/CT calibration.

Dosimetry and theranostics

Patient-specific dosimetry for radiopharmaceutical therapy depends on accurate SPECT quantification of activity over time. Isotopes used in theranostics, such as Lu-177 and I-131, have complex emission spectra, and their accurate quantification is especially sensitive to window placement and scatter correction. An uncorrected or mis-windowed acquisition can bias an absorbed-dose estimate enough to matter for treatment planning 3, 4.

Practical Optimization Tips

1. Peak the camera before each use

Verify photopeak centering as part of daily QC and whenever the radionuclide changes. Energy peaking takes seconds and prevents a day of mis-windowed studies 4.

2. Match the window to the radionuclide and the camera

Do not reuse a Tc-99m window for Tl-201 or Ga-67. Set each photopeak window per the radionuclide's spectrum and the camera's energy resolution, and confirm against the applicable procedure standard 8.

3. Keep window settings constant for serial and quantitative studies

If a patient is followed over time, or a quantitative value will be reported, hold the window and scatter-correction settings fixed across acquisitions. Changing the window changes the scatter fraction and breaks comparability.

4. Enable and validate scatter correction, then leave it consistent

If TEW or a model-based correction is used clinically, confirm the subwindow placement and that the correction is applied in reconstruction. Validate it once against a phantom and keep it consistent 1, 4.

5. Document the settings in the protocol

Record photopeak energies, window widths, subwindow placement, and scatter-correction method in the written acquisition protocol so that every technologist runs the study the same way.

Common pitfalls to avoid

  • Chasing counts by widening the window. More counts at the cost of more scatter is usually a net loss for contrast and quantification.
  • Ignoring photopeak drift. A window that no longer matches the peak quietly wastes counts and degrades uniformity.
  • Copying one radionuclide's window to another. Multi-peak and low-energy emitters need their own settings.
  • Reporting a quantitative value without scatter correction. An SUV from an uncorrected acquisition is not defensible.
  • Changing windows mid-series. Comparability across time depends on constant settings.

Regulatory Considerations

Energy windowing and scatter correction sit inside the instrument quality-control and quantitative-accuracy expectations that govern a nuclear medicine program. The medical use of the byproduct material imaged — Tc-99m, I-131, Ga-67, In-111, Tl-201 — is authorized and overseen under the federal or Agreement State framework for medical use of radioactive material 9.

Key frameworks to reference:

  • NEMA NU 1-2023, Performance Measurements of Gamma Cameras, the standard that defines how energy resolution and related performance parameters are measured and reported, and the edition in force for current acceptance and performance testing 5.
  • IAEA Human Health Series No. 6, Quality Assurance for SPECT Systems, which provides practical guidance on energy windowing, peaking, and routine QC for planar and SPECT acquisition 8.
  • 10 CFR Part 35, Medical Use of Byproduct Material, which governs authorized use, the role of the authorized user, and the radiation safety program for the radionuclides being imaged 9.
  • ICRP Publication 107, Nuclear Decay Data for Dosimetric Calculations, the authoritative source for the photopeak energies on which windows are set 10.

Agreement States administer their own equivalent programs. 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 rules, while Washington, DC and Delaware are regulated directly by the NRC. A facility must confirm which authority issues its license and which QC and recordkeeping requirements apply. DRPS supports these programs through medical physics consulting and PET/CT and nuclear medicine physics services.

Frequently Asked Questions (FAQs)

What is the energy window on a gamma camera?

The energy window is the range of detected photon energies that the camera accepts into the image. It is centered on the radionuclide's photopeak, such as 140 keV for Tc-99m, and set to a width, often 15 to 20 percent, so that most unscattered photopeak photons are recorded while photons that lost energy to Compton scatter are rejected.

Why is a 20 percent window common for Tc-99m?

A symmetric 20 percent window centered on the 140 keV photopeak of Tc-99m spans about 126 to 154 keV. It is a practical compromise: wide enough to capture most unscattered photons given the roughly 9 to 10 percent energy resolution of a sodium iodide gamma camera, but narrow enough to reject a large fraction of lower-energy scattered photons.

What is triple-energy-window scatter correction?

Triple-energy-window (TEW) correction adds two narrow subwindows on either side of the photopeak window. The counts in those subwindows estimate how many scattered photons are contaminating the photopeak window at each pixel, and that estimate is subtracted. It is a widely implemented, pixel-by-pixel scatter correction method.

Does a wider energy window improve image quality?

Not necessarily. A wider window records more counts, reducing statistical noise, but it also admits more scattered photons, which lowers image contrast and degrades quantitative accuracy. Window width is a trade-off between sensitivity and scatter rejection, not a setting to maximize.

How are energy windows set for Ga-67, In-111, and Tl-201?

Radionuclides with multiple photopeaks are imaged with multiple windows. Ga-67 is typically imaged with windows on its 93, 185, and 300 keV peaks; In-111 with windows on 171 and 245 keV; and Tl-201 primarily on its 69 to 83 keV mercury X-rays, often with an added window on 167 keV.

Why does the energy window matter for quantitative SPECT?

Quantitative SPECT converts reconstructed counts into an activity concentration. Scattered photons counted in the photopeak window inflate those counts and bias the result, so a correctly centered window plus a validated scatter correction is a prerequisite for accurate quantification, which for Tc-99m can reach about plus or minus 10 percent with CT-based corrections.

How often should energy peaking be checked?

Photopeak centering should be verified before each use day as part of routine quality control, and whenever a different radionuclide is imaged. Drift in the photopeak position relative to the window is a common, correctable cause of lost counts, reduced uniformity, and degraded images.

Key Takeaways

  • The energy window is the camera's first scatter rejector. It accepts photopeak photons and discards the lower-energy scatter that falls outside it.
  • Window width is a trade-off. Wider means more counts but more scatter; narrower means cleaner but noisier. The 15 to 20 percent photopeak window for Tc-99m balances the two.
  • Energy resolution sets what is possible. A sharper photopeak, as on CZT cameras, allows a narrower window that rejects more scatter for the same yield.
  • Multi-peak and low-energy radionuclides need their own settings. Ga-67, In-111, and Tl-201 are not imaged with a Tc-99m window.
  • Triple-energy-window correction turns subwindows into a scatter estimate that is subtracted pixel by pixel, supporting quantitative SPECT.
  • Quantitative accuracy depends on it. Tc-99m SPECT can reach about plus or minus 10 percent with proper windowing, scatter correction, and CT-based attenuation correction.

Conclusion

The energy window looks like a small number on an acquisition screen, but it governs how much scatter enters the image, how much contrast survives, and whether a quantitative value can be trusted. Set it on the photopeak, size it to the camera's energy resolution and the radionuclide's spectrum, add scatter subwindows where correction is used, and keep the settings consistent for serial and quantitative work.

A department that treats energy peaking as a daily checkpoint and documents its window and scatter-correction settings in the written protocol will produce more consistent images, cleaner quantification, and a program that is easier to defend during review. The medical physicist's role is to validate those settings, confirm the scatter correction performs as expected, and connect the windowing choices to the camera's measured energy resolution.

How DRPS Can Help

Diagnostic Radiation Physics Services helps nuclear medicine and PET/CT facilities turn instrument physics into practical, documented workflows. For gamma cameras and SPECT/CT, this can include acceptance and performance testing, energy-resolution and uniformity QC, scatter- and attenuation-correction validation, quantitative SPECT calibration, and acquisition-protocol review aligned with NEMA, IAEA, and SNMMI guidance.

DRPS supports facilities across our service locations, including Florida, Maryland, Virginia, Washington DC, California, Nevada, New York, Pennsylvania, New Jersey, and Delaware, through PET/CT and nuclear medicine physics and medical physics consulting.

Related Resources

References

  1. Ogawa K, Harata Y, Ichihara T, Kubo A, Hashimoto S. A practical method for position-dependent Compton-scatter correction in single photon emission CT. IEEE Trans Med Imaging. 1991;10(3):408-412. doi:10.1109/42.97591. PubMed
  2. Hutton BF, Buvat I, Beekman FJ. Review and current status of SPECT scatter correction. Phys Med Biol. 2011;56(14):R85-R112. doi:10.1088/0031-9155/56/14/R01. PubMed
  3. Bailey DL, Willowson KP. An evidence-based review of quantitative SPECT imaging and potential clinical applications. J Nucl Med. 2013;54(1):83-89. doi:10.2967/jnumed.112.111476. PubMed
  4. Zanzonico P. Routine quality control of clinical nuclear medicine instrumentation: a brief review. J Nucl Med. 2008;49(7):1114-1131. doi:10.2967/jnumed.107.050203. PubMed
  5. National Electrical Manufacturers Association. NEMA NU 1-2023: Performance Measurements of Gamma Cameras. 2023. nema.org
  6. Bai J, Hashimoto J, Ogawa K, Nakahara T, Suzuki T, Kubo A. Scatter correction based on an artificial neural network for 99mTc and 123I dual-isotope SPECT in myocardial and brain imaging. Ann Nucl Med. 2007;21(1):25-32. doi:10.1007/BF03033996. PubMed
  7. Ben-Haim S, Kennedy J, Keidar Z. Novel cadmium zinc telluride devices for myocardial perfusion imaging — technological aspects and clinical applications. Semin Nucl Med. 2016;46(4):273-285. doi:10.1053/j.semnuclmed.2016.01.002. PubMed
  8. International Atomic Energy Agency. Quality Assurance for SPECT Systems (IAEA Human Health Series No. 6). Vienna: IAEA; 2009. iaea.org
  9. U.S. Nuclear Regulatory Commission. 10 CFR Part 35: Medical Use of Byproduct Material. ecfr.gov
  10. International Commission on Radiological Protection. ICRP Publication 107: Nuclear Decay Data for Dosimetric Calculations. Ann ICRP. 2008;38(3). icrp.org