Quantitative SPECT/CT: Calibration & SUV

Quantitative SPECT/CT turns a SPECT scan from a map of relative counts into an absolute measurement of activity concentration — becquerels per milliliter in every voxel — which is what makes SUV in SPECT and patient-specific dosimetry possible. Achieving it requires a traceable system calibration factor, CT-based attenuation correction, scatter correction, collimator–detector resolution recovery, and partial-volume correction, all held stable by QC. Without that chain, vendor reconstructions of the same phantom can disagree by more than 100%. ¹²⁵

Introduction

For most of its history, SPECT has been read qualitatively: brighter means more uptake, and the clinician compares regions within a single scan. That is adequate for many diagnostic questions, but it cannot answer the quantitative ones that modern nuclear medicine increasingly asks. How much activity is in this lesion? What absorbed dose did this kidney receive from a therapy administration? Did uptake change between cycles by a clinically meaningful amount? Answering those requires an absolute number, not a relative one. ¹²

Quantitative SPECT/CT provides that number. By combining the emission data with the co-registered CT — which supplies the attenuation map and anatomic boundaries — and by anchoring the reconstruction to a measured system calibration factor, a modern SPECT/CT system can report activity concentration in Bq/mL and, from there, a standardized uptake value (SUV) directly analogous to PET. This capability underpins the dosimetry that theranostics programs such as Lu-177 DOTATATE and Lu-177 PSMA depend on. ²³

But quantitative SPECT is unforgiving. Every correction in the chain must be right, the calibration must be traceable, and the QC must confirm the system has not drifted — because a multicenter study of state-of-the-art systems found that quantified activity varied by more than a factor of two between vendors when each used its own default reconstruction. ⁵ This article walks through the calibration factor, the correction chain, the SUV calculation, the recovery-coefficient concept, and the QC that keeps it all honest. DRPS supports quantitative SPECT/CT programs 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

From counts to becquerels

In a conventional SPECT reconstruction, each voxel holds a count value with no absolute physical meaning; in a quantitative SPECT reconstruction, each voxel holds an activity concentration in Bq/mL. The bridge between the two is a system calibration factor (also called the system sensitivity or volume sensitivity), which encodes how many counts per second the camera registers per unit of activity in the field of view under a defined acquisition and reconstruction protocol. ¹⁶

The calibration factor is not a single universal constant. It depends on the radionuclide, the collimator, the energy windows, the acquisition geometry, and — critically — the reconstruction algorithm and its settings. A calibration is valid only for the exact protocol under which it was measured. Change the collimator, the energy window, or the number of iterations, and the calibration must be re-established. ¹²

The corrections that make quantification possible

Absolute quantification is only meaningful if the physical degradations that corrupt SPECT counts are corrected. The essential corrections are:

Attenuation correction (AC): photons are absorbed and scattered on the way out of the body. The co-registered CT provides a patient-specific attenuation map, converted to the emission energy, so counts can be corrected for the tissue they traversed. ¹
Scatter correction (SC): photons that scatter can be mis-positioned. Energy-window–based methods (for example, an adjacent lower scatter window) or model-based scatter estimation remove this bias. ¹⁵
Resolution recovery / collimator–detector response (CDR) modeling: the collimator blurs the image in a distance-dependent way; modeling the CDR in the reconstruction partially restores resolution. ¹
Partial-volume correction (PVC): structures smaller than roughly 2–3 times the system resolution do not recover their true concentration; recovery coefficients correct for this. ¹⁵
Dead-time and decay corrections: at the high count rates of some therapy imaging, dead-time losses must be corrected, and all activity is decay-corrected to a reference time. ²³

For the SPECT/CT hardware QC that keeps registration and uniformity valid — prerequisites for trustworthy quantification — see our guide to SPECT/CT quality control.

Vendor implementations

Commercial quantitative SPECT reconstructions include Siemens xSPECT Quant and Broad Quantification, GE HealthCare Evolution with Q.Metrix, and vendor-neutral options such as Hermes SUV SPECT. Each implements the correction chain, but the specific scatter model, resolution-recovery approach, and default settings differ — which is exactly why the same phantom can quantify differently across systems unless the protocol is standardized. ⁵ Standardization to a traceable calibration is the theme that runs through the MIRD and EANM guidance.

Key Technical Principles

The calibration factor

The calibration factor $C_{F}$ relates a measured count rate to a known activity. In a calibration acquisition of a source with accurately known activity $A_{cal}$ (in MBq) that yields a total count rate $R_{cal}$ (in counts per second), the calibration factor is:

C_{F} = \frac{R _{cal}}{A _{cal}} [\frac{cps}{MBq}]

The known activity must be cross-calibrated to a dose calibrator whose accuracy is itself traceable to a national standard such as NIST — the calibration is only as good as the activity measurement behind it. For the dose-calibrator QC that anchors this traceability, see our dose calibrator quality control guide.

Once $C_{F}$ is known, the activity $A_{VOI}$ in a patient volume of interest containing $N$ counts acquired over time $t$ (seconds) is:

A_{VOI} = \frac{N / t}{C _{F}} [MBq]

and the activity concentration is $c = A_{VOI} / V_{VOI}$ , with $V_{VOI}$ the volume of the region (in mL).

The standardized uptake value in SPECT

Once activity concentration is available, the SUV follows exactly as in PET. Normalizing the tissue activity concentration by the administered activity distributed over body mass:

SUV = \frac{c _{tissue} [ kBq/mL ]}{A _{administered} [ kBq ] / m _{body} [ g ]}

assuming a tissue density of approximately 1 g/mL so that mL and g are interchangeable.

Worked example. A patient weighing 70 kg is administered 740 MBq of Tc-99m MDP for a quantitative bone SPECT/CT. A calibrated VOI over a lesion reports an activity concentration of 45 kBq/mL. The administered activity per unit mass is:

\frac{A _{administered}}{m _{body}} = \frac{740 , 000 kBq}{70 , 000 g} \approx 10.6 \frac{kBq}{g}

so the lesion SUV is:

SUV = \frac{45}{10.6} \approx 4.2

(decay-correcting the administered activity to the acquisition time in an actual calculation). An SUV of about 4 is a concrete, comparable number — it can be trended across follow-up scans in a way that "the lesion looks hotter" never could.

Recovery coefficients and the partial-volume effect

Because SPECT resolution is limited (on the order of a centimeter after reconstruction), small objects lose apparent concentration. The recovery coefficient (RC) quantifies the loss:

RC = \frac{c _{measured}}{c _{true}}

RC approaches 1 for large objects and falls well below 1 for small ones. In a multicenter phantom study using spheres from 0.5 to 113 mL at a 10:1 sphere-to-background ratio, recovery coefficients decreased with decreasing sphere volume, and the inter-system variation in RC ranged from 0.06 to 0.41 (up to about a 118% quantification difference) using vendor-specific reconstructions, improving to 0.02–0.19 (about 38%) with a standardized vendor-neutral reconstruction. ⁵ Partial-volume correction divides the measured concentration by an object-appropriate RC — but the RC curve is protocol- and system-specific, so it must be measured, not assumed.

The correction chain at a glance

Correction	Physical problem addressed	Typical method	Consequence if omitted
Attenuation (AC)	Photon absorption in tissue	CT-based attenuation map at emission energy	Large, depth-dependent underestimation
Scatter (SC)	Mis-positioned scattered photons	Energy-window (e.g. adjacent lower window) or model-based	Overestimation, poor contrast
Resolution recovery (CDR)	Distance-dependent collimator blur	Modeling collimator–detector response in reconstruction	Reduced resolution and contrast recovery
Partial-volume (PVC)	Small-object signal loss	Recovery coefficients from sphere phantom	Underestimation of small lesions
Dead-time	Count losses at high rates	Count-rate–dependent correction	Underestimation at high activity
Decay	Physical decay during/between scans	Decay-correct to reference time	Time-point errors in dosimetry

Clinical Impact

Quantitative SPECT/CT changes what SPECT can be used for. Three areas show the impact most clearly.

First, theranostics dosimetry. Patient-specific absorbed-dose calculation for Lu-177 and other therapies requires an accurate activity estimate at each imaging time point, from which a time–activity curve and cumulated activity are derived. Quantitative SPECT/CT provides those estimates; the MIRD 23 framework and the Lu-177–specific MIRD 26 / EANM guidance exist precisely to standardize them. This ties directly to our Lu-177 theranostics dosimetry and MIRD schema internal dosimetry discussions. ²³

Second, quantitative diagnostic SPECT. Bone SPECT/CT with SUV is the most mature diagnostic application, allowing lesion uptake to be reported as a number and compared over time — bringing SPECT closer to the quantitative footing PET has long enjoyed, discussed in our PET SUV quantification article.

Third, multicenter and longitudinal consistency. Because uncontrolled vendor differences can exceed 100%, quantitative accuracy is only clinically usable when acquisition and reconstruction are standardized and the calibration is traceable. A quantified SUV that cannot be reproduced on the next scanner, or at the next time point, is not a biomarker — it is noise with a decimal point. ⁵

Practical Optimization Tips

A defensible quantitative SPECT/CT program follows a disciplined setup and QC routine. The following reflects common practice, adapted to the system and the current guidance.

1. Establish the calibration correctly

Measure the calibration factor with a source of accurately known activity, using a dose calibrator whose accuracy and linearity are under a current QC program and traceable to a national standard.
Acquire and reconstruct the calibration source with the exact clinical protocol — same collimator, energy windows, matrix, iterations, and corrections.
Document the calibration factor, the conditions, and the date; re-verify periodically and after any relevant hardware or software change. ¹⁶

2. Standardize acquisition and reconstruction

Fix the energy windows (for Lu-177, the 208 keV photopeak with an appropriate scatter window; for Tc-99m, the 140 keV photopeak), collimator selection, orbit, and projection sampling.
Lock the reconstruction settings — iterations, subsets, filters, and which corrections are enabled — and change them only through a controlled process, because they alter the calibration. ⁵

3. Characterize recovery coefficients

Measure an RC curve from a sphere phantom spanning clinically relevant volumes under the clinical protocol.
Apply object-appropriate partial-volume correction for small-lesion quantification, and state the uncertainty for very small volumes. ⁵

4. Verify with QC before trusting a number

Confirm SPECT/CT registration, uniformity, and center-of-rotation are within tolerance, since misregistration corrupts attenuation correction.
Periodically re-image a known-activity phantom and confirm the recovered concentration is within your accuracy goal before reporting quantitative results clinically.

Common pitfalls to avoid

Reusing a calibration across protocols. A calibration factor is protocol-specific; applying it to a different collimator, window, or reconstruction invalidates the result.
Ignoring partial-volume effects. Small lesions are systematically underestimated without RC correction.
Untraceable activity. If the dose calibrator behind the calibration is not accurate and traceable, every downstream Bq/mL is wrong by the same factor.
Assuming vendor equivalence. Default reconstructions differ; standardize before comparing across systems or sites. ⁵
Skipping registration QC. A CT-to-SPECT misalignment produces attenuation-correction artifacts that masquerade as real uptake changes.

Regulatory Considerations

Quantitative SPECT/CT is governed less by a single regulation than by the professional and consensus standards that define correct practice, together with the radioactive-material rules that apply to the underlying therapy program.

MIRD Pamphlet No. 23 provides the general framework for quantitative SPECT for patient-specific dosimetry, and MIRD Pamphlet No. 24 (I-131) and MIRD Pamphlet No. 26 (Lu-177, joint EANM/MIRD) give radionuclide-specific quantification guidance. ¹³⁷
IAEA Human Health Reports No. 9 consolidates the concepts, requirements, and methods of quantitative nuclear medicine imaging, including calibration and cross-calibration practice. ⁶
NEMA NU-1 defines standardized performance measurements (including system sensitivity) for gamma cameras, complementing the calibration workflow; see our gamma camera NEMA NU-1 performance testing guide. ⁸
10 CFR Part 35 (or the Agreement State equivalent) governs the medical use of the byproduct material being imaged, including written directives and the authorized-user framework for therapy; quantitative imaging supports the dosimetry that a growing number of therapy programs perform. Radioactive material is regulated by the NRC or Agreement States, distinct from the FDA/state framework for x-ray machines. Among the states DRPS serves, Florida, Maryland, Virginia, California, Nevada, Pennsylvania, New York, and New Jersey are Agreement States, while Washington DC and Delaware are regulated directly by the NRC. ⁴

Documenting the calibration, the corrections, the recovery-coefficient characterization, and the QC is what makes a quantitative result defensible for clinical or dosimetric use. For programs building toward therapy dosimetry, this should be coordinated with medical physicist consulting and the facility's radioactive-material license.

Frequently Asked Questions (FAQs)

What is quantitative SPECT/CT?

Quantitative SPECT/CT is SPECT imaging that reports an absolute activity concentration — becquerels per milliliter — in each voxel rather than only relative counts. It relies on a traceable system calibration factor plus CT-based attenuation correction, scatter correction, resolution recovery, and partial-volume correction, so the reconstructed value can be compared across patients, time points, and systems.

What is the SPECT calibration factor?

The calibration (or sensitivity) factor converts the count rate a camera detects into activity, typically expressed in counts per second per megabecquerel (cps/MBq). It is measured by scanning a source of accurately known activity — cross-calibrated to a NIST-traceable dose calibrator — under the same acquisition and reconstruction settings used clinically.

Can SPECT report an SUV like PET?

Yes. Once a SPECT/CT system is quantitatively calibrated, the voxel activity concentration can be normalized by administered activity and body mass to yield a standardized uptake value (SUV), analogous to PET. This is increasingly used in bone SPECT/CT and in therapy response assessment, provided the calibration and corrections are valid.

Why do quantitative SPECT results differ between vendors?

A multicenter study found recovery coefficients — and therefore quantified activity — varied substantially between systems using vendor-specific reconstructions, with differences up to about 118% for small volumes, improving to about 38% when a standardized, vendor-neutral reconstruction was applied. Differences in scatter modeling, resolution recovery, and reconstruction settings drive the spread, which is why standardization matters for multicenter dosimetry.

What is a recovery coefficient?

A recovery coefficient is the imaged activity concentration divided by the true activity concentration for an object of known size. Because of limited spatial resolution, small structures recover less than their true value (the partial-volume effect), so recovery coefficients — usually measured from a sphere phantom — are used to correct small-lesion quantification.

Why is quantitative SPECT/CT important for radiopharmaceutical therapy?

Patient-specific dosimetry for therapies such as Lu-177 requires accurate activity estimates at each imaging time point. Quantitative SPECT/CT provides the absolute activity concentration that feeds the time–activity curve and the absorbed-dose calculation, which is why MIRD and EANM guidelines specify quantitative SPECT methods for dosimetry.

How is a quantitative SPECT/CT system kept in calibration?

The system calibration factor is established with a known-activity phantom and periodically re-verified, the dose calibrator used for cross-calibration is maintained under its own QC program, and routine uniformity, center-of-rotation, and SPECT/CT registration checks confirm the corrections remain valid. Drift in any of these degrades quantitative accuracy.

Key Takeaways

Quantification means Bq/mL, not counts. A system calibration factor plus attenuation, scatter, resolution, partial-volume, dead-time, and decay corrections turn a SPECT scan into an absolute measurement. ¹
Calibration must be traceable and protocol-specific. The calibration factor is only valid for the exact acquisition and reconstruction it was measured under, and only as accurate as the NIST-traceable dose calibrator behind it. ⁶
SUV comes for free once calibrated. Normalizing activity concentration by administered activity and body mass yields an SUV directly analogous to PET.
Small objects need recovery-coefficient correction. The partial-volume effect systematically underestimates small lesions unless RC-based PVC is applied. ⁵
Standardization is essential. Vendor-default reconstructions can disagree by more than 100%; standardized protocols and calibrations shrink that dramatically. ⁵
Dosimetry depends on it. MIRD 23/24/26 and EANM guidance specify quantitative SPECT because absorbed-dose accuracy hinges on activity accuracy. ¹³

Conclusion

Quantitative SPECT/CT is one of the most consequential capabilities in modern nuclear medicine physics, because it converts a historically qualitative modality into a measurement instrument. That conversion is not automatic: it depends on a traceable calibration factor, a complete and validated correction chain, characterized recovery coefficients, and QC that confirms the system has not drifted. The reward is substantial — SUV-based diagnostic reporting and the patient-specific dosimetry that theranostics requires — but the discipline is non-negotiable, because uncontrolled vendor and protocol differences can change the answer by more than a factor of two. A qualified medical physicist establishes the calibration, standardizes the protocol, characterizes the corrections, and maintains the QC that lets a clinician trust the number. ¹²⁵

How DRPS Can Help

Diagnostic Radiation Physics Services supports nuclear medicine and theranostics programs with quantitative SPECT/CT calibration, cross-calibration to traceable dose-calibrator standards, recovery-coefficient characterization, protocol standardization, and the SPECT/CT QC that keeps quantification valid. Our board-certified medical physicists provide PET/CT and nuclear medicine physics and medical physicist consulting across our service locations, including Florida, Maryland, Virginia, Washington DC, California, Nevada, Pennsylvania, New York, New Jersey, and Delaware.

A quantitative SPECT/CT program is only as trustworthy as its calibration and QC — that is where an experienced physicist adds the most value.

Related Resources

References

Dewaraja YK, Frey EC, Sgouros G, et al. MIRD pamphlet No. 23: quantitative SPECT for patient-specific 3-dimensional dosimetry in internal radionuclide therapy. J Nucl Med. 2012;53(8):1310-1325. doi:10.2967/jnumed.111.100123. doi.org
Willowson K, Bailey DL, Baldock C. Quantitative SPECT reconstruction using CT-derived corrections. Phys Med Biol. 2008;53(12):3099-3112. doi:10.1088/0031-9155/53/12/002. doi.org
Ljungberg M, Celler A, Konijnenberg MW, et al. MIRD Pamphlet No. 26: Joint EANM/MIRD guidelines for quantitative 177Lu SPECT applied for dosimetry of radiopharmaceutical therapy. J Nucl Med. 2016;57(1):151-162. doi:10.2967/jnumed.115.159012. doi.org
U.S. Nuclear Regulatory Commission. 10 CFR Part 35: Medical Use of Byproduct Material. ecfr.gov
Peters SMB, Meyer Viol SL, van der Werf NR, et al. Variability in lutetium-177 SPECT quantification between different state-of-the-art SPECT/CT systems. EJNMMI Phys. 2020;7(1):9. doi:10.1186/s40658-020-0278-3. doi.org
International Atomic Energy Agency. Quantitative Nuclear Medicine Imaging: Concepts, Requirements and Methods. IAEA Human Health Reports No. 9. Vienna: IAEA; 2014. iaea.org
Dewaraja YK, Ljungberg M, Green AJ, et al. MIRD pamphlet No. 24: guidelines for quantitative 131I SPECT in dosimetry applications. J Nucl Med. 2013;54(12):2182-2188. doi:10.2967/jnumed.113.122390. doi.org
National Electrical Manufacturers Association. NEMA NU 1: Performance Measurements of Gamma Cameras. Rosslyn, VA: NEMA. nema.org