Quantitative Bone SPECT/CT with SUV
Quantitative bone SPECT/CT turns a bone scan into a measurement: with CT-based attenuation and scatter correction, resolution recovery, and a cross-calibrated counts-to-activity factor, the system reports standardized uptake values (SUV) in grams per milliliter instead of relative brightness. Those numbers help separate degenerative from metastatic uptake and support serial follow-up, but they are platform-, tracer-, and method-specific and only as trustworthy as the calibration and QC behind them. 123
For decades the planar and SPECT bone scan was read qualitatively: a "hot spot" was more or less intense than its surroundings. Modern SPECT/CT systems can instead convert reconstructed counts into an absolute activity concentration, and from there into SUV, bringing to bone SPECT the quantitative language long used in PET. 34 This article explains how quantitative bone SPECT/CT works, the calibration chain and corrections it depends on, a worked SUV calculation, what the SUV numbers mean clinically, the QC that keeps them honest, and how the program fits into accreditation. DRPS supports this work through its PET/CT and nuclear medicine physics and medical physics consulting services. For the general methodology that underpins all of this, see our companion guide to quantitative SPECT/CT calibration and SUV.
Introduction
The defining idea of quantitative bone SPECT/CT is cross-calibration: making the number of reconstructed counts in a voxel correspond to a known activity concentration, so the image carries physical units. A conventional SPECT reconstruction produces counts that depend on camera sensitivity, attenuation, scatter, and reconstruction settings. Quantitative SPECT/CT removes those dependencies as far as practical and anchors the result to a measured standard. 356
Bone scintigraphy itself remains one of the most common nuclear medicine studies, using technetium-99m–labeled diphosphonates such as MDP, HDP, or DPD that adsorb onto sites of active bone turnover. 1 The clinical question is frequently whether increased uptake reflects benign degenerative change, trauma, or metastatic disease — a question where a defensible number can add specificity to the qualitative read. 112 Quantitative SPECT/CT, validated first in other applications, now joins PET as a modality that can report activity concentration and SUV. 4 For how the underlying corrections are applied in practice, see our guides to SPECT/CT attenuation correction and the PET SUV concept that bone SPECT SUV mirrors.
Topic Explanation
What is quantitative bone SPECT/CT?
Quantitative bone SPECT/CT is a bone SPECT acquisition reconstructed with CT-based attenuation correction, scatter correction, and collimator–detector resolution recovery, then cross-calibrated so reconstructed counts become an activity concentration in kilobecquerels per milliliter (kBq/mL). Normalized to administered activity and body mass, that concentration becomes the SUV. 236
The chain from a raw scan to an SUV has four links:
- Reconstruction with corrections. Iterative reconstruction (ordered-subset expectation maximization) incorporating CT-based attenuation correction, scatter correction, and 3D resolution recovery. 56
- System cross-calibration. A uniform phantom of known activity concentration, measured in a dose calibrator, is scanned to derive a counts-to-kBq/mL sensitivity factor. 26
- Activity-concentration image. The corrected, calibrated reconstruction reports kBq/mL per voxel.
- SUV normalization. The VOI activity concentration is normalized to administered activity per body mass. 23
The European Association of Nuclear Medicine has published a practice guideline specifically for quantitative SPECT/CT that lays out these requirements, alongside its long-standing bone scintigraphy guideline. 12
SUV metrics and what they mean
As in PET, several SUV metrics coexist and are not interchangeable:
- SUVmax — the single hottest voxel in the VOI; simple and reproducible but noise-sensitive.
- SUVpeak — the mean of a small fixed sub-volume around the hottest region; more stable than SUVmax.
- SUVmean — the average over the whole VOI; depends on how the VOI is drawn.
Every reported SUV must state which metric, which tracer, which reconstruction, and which site it refers to, because those choices change the number materially. 813
Key Technical Principles
From counts to activity concentration
A gamma camera records counts, not activity. The confounders that stand between counts and a true activity concentration are, explicitly, photon attenuation, photon scatter, the partial volume effect, and patient motion. 7 Quantitative SPECT/CT addresses the first two with CT-based attenuation correction and model-based scatter correction, improves spatial response with resolution recovery, and depends on the user to limit motion and to account for partial-volume losses in small lesions. 567
The system calibration factor links corrected counts to activity concentration. In practice a uniform cylinder filled with a measured activity concentration (traceable to the dose calibrator) is scanned and reconstructed with the clinical settings; the ratio of known kBq/mL to reconstructed counts is the sensitivity calibration. 26 With this chain in place, published work shows absolute quantitative accuracy for technetium-99m within about ±10% of the true concentration in vivo. 3 A method using CT-derived corrections reported a mean difference between calculated and known activity of about −1% with a range of roughly ±7% across studies, 5 and an OSEM-3D implementation with CT attenuation and scatter correction — the basis of a commercial quantitative bone platform — reported phantom accuracy within 3.6% and patient accuracy within 1.1% against a urine-activity reference. 6
The SUV definition
Standardized uptake value normalizes the measured VOI activity concentration to the administered activity per unit body mass:
where
Worked SUV example
Consider a lumbar vertebral focus imaged with technetium-99m MDP. The patient received 740 MBq, and SPECT/CT was acquired 3 hours after injection. Technetium-99m has a physical half-life of about 6.0 hours, so the decay factor over 3 hours is:
The activity remaining at scan time is therefore:
For a 70 kg patient (
That value falls within the range reported for bone metastases in quantitative SPECT/CT series (for example, metastatic SUVmax of about 24.8 ± 16.3 in one differential-diagnosis study), which is well above typical degenerative values. 13 The example also shows why decay correction and an accurate body mass are not optional: an error in either propagates directly into the SUV.
Reproducibility
A number is only useful if it repeats. A test–retest study of quantitative bone SPECT/CT in prostate-cancer patients, scanning each twice a few days apart, reported an intraclass correlation coefficient of about 0.97 for both SUVmax and SUVpeak, with a repeatability coefficient near 31% for SUVmax and 28% for SUVpeak, and a mean absolute percentage difference around 12%. 8 The practical reading: a change smaller than roughly one-third of the value may be within measurement noise, which sets a floor on what can be called a real change on follow-up. Normal-vertebra studies likewise show large variation between patients but small variation within a patient, with the body-weight–based maximum SUV giving the lowest coefficient of variation. 9
Clinical Impact
The clearest value of quantitative bone SPECT/CT is adding specificity to the benign-versus-malignant question that bone scans are so often asked. Across several single-center series, metastatic lesions show higher SUV than degenerative or benign lesions, and an SUV cutoff improves specificity relative to the qualitative read — but the cutoffs are study-specific, as the table makes plain.
| Study context | Tracer / platform | SUV metric & cutoff | Reported performance |
|---|---|---|---|
| Prostate metastases vs. osteoarthritis 12 | Tc-99m DPD, xSPECT | SUV cutoff ≈ 19.5 g/mL | Sensitivity 87%, specificity 92% |
| Metastatic vs. benign hotspots 13 | Tc-99m, quantitative SPECT/CT | SUVmax cutoff ≈ 17.7 | AUC ≈ 0.69 |
| Benign vs. malignant bone lesions 14 | Tc-99m, MLEM quantitative SPECT/CT | SUVmax cutoff ≈ 26.6 g/mL | Specificity 100%, sensitivity 65% |
| Benign vs. malignant bone lesions 14 | Tc-99m, MLEM quantitative SPECT/CT | SUVmean cutoff ≈ 12.8 g/mL | Sensitivity 78%, specificity 94% |
The cutoffs differ by roughly 40–50% across these studies, driven by differences in tracer, reconstruction algorithm and platform, SUV metric, and anatomic site — which is exactly why a single universal threshold does not exist and why each number must be reported with its full context. 121314 For reference, non-lesion bone in one series showed a mean SUVmax near 5.9. 13
Two further clinical points matter. First, quantitative bone SPECT/CT SUVs correlate strongly with fluorine-18 sodium fluoride PET SUVs in bone metastases (correlation coefficients at or above 0.80), though the PET values run higher — by a median of about 21% for SUVmax — so SPECT and PET numbers are related but not identical. 11 Lesion-to-normal SUV ratios agreed more closely between the two modalities than the absolute values did. 11 Second, because SUV is reproducible within roughly a third, it opens the door to serial, objective follow-up of known lesions, provided the acquisition and reconstruction are held constant between scans. 8
Practical Optimization Tips
Cross-calibrate, and keep it current
The counts-to-activity factor is the foundation. Perform the cross-calibration with a uniform phantom traceable to the dose calibrator, using the exact clinical reconstruction settings, and recalibrate on the schedule the physicist sets so the factor tracks camera sensitivity over time. 26 A drifting dose calibrator or camera sensitivity silently biases every SUV.
Freeze the acquisition and reconstruction for serial work
SUV depends on acquisition time, matrix, iterations, subsets, and correction settings. When a lesion is followed over time, hold all of these constant; a reconstruction change between scans can mimic or mask a real SUV change. One phantom study showed that a bone-specific reconstruction kept SUV close to the true value and that acquisition time could be shortened substantially with comparable SUV, but such changes should be validated before they enter a serial protocol. 10
Mind the partial volume effect
Small lesions lose apparent activity concentration because their size approaches the system resolution, biasing SUVmax and SUVmean downward. 7 For sub-centimeter foci, interpret SUV cautiously and consider SUVpeak, which is more stable, while recognizing partial-volume losses remain. 78
Report the metric, tracer, platform, and site
Because cutoffs are not interchangeable, every reported SUV should carry its context: SUVmax vs. SUVpeak vs. SUVmean, the radiopharmaceutical, the reconstruction platform, and the anatomic site. 121314 A threshold lifted from a paper using a different tracer or reconstruction can be misleading.
Common pitfalls to avoid
- Treating SUV as calibration-free. Without cross-calibration and CT-based corrections, the image is not quantitative. 36
- Applying one scanner's cutoff everywhere. Cutoffs are tracer-, platform-, metric-, and site-specific. 121314
- Over-reading small changes on follow-up. A change within roughly the repeatability coefficient may be noise. 8
- Ignoring the partial volume effect. Small lesions read low. 7
- Changing reconstruction mid-series. It breaks comparability. 10
Regulatory Considerations
Quantitative bone SPECT/CT sits within the facility's nuclear medicine QC program, accreditation obligations, and radioactive material license, and the quantitative layer adds calibration and documentation requirements on top of routine camera QC. The program should be documented and overseen by a qualified medical physicist so SUV results are defensible.
Key frameworks to reference:
- EANM practice guideline for quantitative SPECT/CT — the methodology reference for calibration, corrections, and reporting of absolute SPECT quantification. 2
- EANM practice guidelines for bone scintigraphy — the clinical procedure standard for bone imaging into which quantification is integrated. 1
- NEMA NU-1 performance methodology — the standard framework for characterizing gamma-camera performance (sensitivity, uniformity, resolution) that underpins a camera's fitness for quantitative work.
- 10 CFR Part 35 / Agreement State medical-use rules — govern the radioactive material license and the facility's required QC posture; of the states DRPS serves, Florida, Maryland, Virginia, California, Nevada, Pennsylvania, New York, and New Jersey are NRC Agreement States, while Washington, DC and Delaware are regulated directly by the NRC.
Routine gamma-camera QC — uniformity, center-of-rotation, energy peaking — remains the prerequisite for any quantitative claim, because a camera that fails basic performance cannot produce trustworthy SUVs. For the broader quantitative methodology and the day-to-day camera QC that supports it, see quantitative SPECT/CT calibration and SUV and nuclear medicine diagnostic reference levels.
Frequently Asked Questions (FAQs)
What is quantitative bone SPECT/CT?
It is a bone SPECT acquisition reconstructed with CT-based attenuation and scatter correction and resolution recovery, then cross-calibrated so reconstructed counts become an activity concentration in kBq/mL. Normalized to administered activity and body mass, that concentration is the SUV, in g/mL. 236
How is SUV calculated in bone SPECT?
SUV is the measured VOI activity concentration divided by the administered activity (decay-corrected to scan time) per unit body mass, treating 1 mL of tissue as about 1 g. SUVmax is the hottest voxel, SUVpeak a small fixed sub-volume, and SUVmean the VOI average; the type must be stated. 238
Can SUV distinguish a degenerative lesion from a metastasis?
Quantitative SUV adds specificity: metastatic lesions tend to show higher SUVmax than benign or degenerative lesions, with cutoffs that improve specificity. But the cutoffs are tracer-, platform-, metric-, and site-specific and are not interchangeable, so they support rather than replace clinical and anatomic correlation. 121314
How reproducible are bone SPECT SUV measurements?
A test–retest study reported an intraclass correlation coefficient near 0.97 for SUVmax and SUVpeak and a repeatability coefficient of about 31% for SUVmax, meaning a change smaller than roughly a third may be within measurement noise. 8
Why does bone SPECT SUV need a calibration factor?
A camera records counts, not activity. Cross-calibration against a dose calibrator using a uniform phantom, after attenuation, scatter, and resolution-recovery corrections, converts counts to kBq/mL. Without it, the image is not quantitative. 26
Key Takeaways
- Quantitative bone SPECT/CT reports SUV in g/mL. CT-based correction plus cross-calibration converts counts into an absolute activity concentration. 236
- Accuracy is achievable but earned. Technetium-99m quantification within about ±10% is reported, contingent on calibration and corrections. 356
- SUV adds specificity to benign vs. malignant. Metastases read higher, and cutoffs improve specificity — but they are study-specific. 121314
- Reproducibility sets the follow-up floor. ICC near 0.97 with a repeatability coefficient near 31% means sub-third changes may be noise. 8
- Context is mandatory. Report the SUV metric, tracer, reconstruction platform, and site; cutoffs do not transfer. 121314
- Camera QC comes first. Uniformity, center-of-rotation, and energy QC underpin any quantitative claim. 2
Conclusion
Quantitative bone SPECT/CT upgrades the bone scan from a picture to a measurement, and that measurement — SUV in g/mL — can sharpen the benign-versus-malignant judgment and enable objective follow-up. The caveats are as important as the capability: the number is only as good as the cross-calibration and corrections behind it, the reproducibility limit means small serial changes may be noise, and the published cutoffs are tied to a specific tracer, platform, SUV metric, and site. Programs that treat quantification as a disciplined, documented, physicist-supervised process — not a button on the console — are the ones whose SUVs can be trusted. 268
How DRPS Can Help
Diagnostic Radiation Physics Services helps nuclear medicine facilities build defensible quantitative imaging programs. For quantitative bone SPECT/CT, this can include designing and verifying the system cross-calibration, selecting attenuation, scatter, and resolution-recovery reconstruction settings, establishing recalibration and QC schedules and normal ranges, and providing PET/CT and nuclear medicine physics, accreditation support, and medical physics consulting performed by board-certified medical physicists.
DRPS supports facilities across our service locations, including Florida, Maryland, Virginia, Washington DC, California, Nevada, New York, Pennsylvania, New Jersey, and Delaware.
A trustworthy SUV is not a number off the console. It is the end of a calibrated, corrected, and documented chain the whole clinical team can rely on.
Related Resources
- Quantitative SPECT/CT calibration and SUV
- SPECT/CT attenuation correction
- PET SUV quantification
- Nuclear medicine diagnostic reference levels
- Voxel-based dosimetry for radiopharmaceutical therapy
- PET/CT and nuclear medicine physics
- Medical physicist consulting
- Accreditation support
References
- Van den Wyngaert T, Strobel K, Kampen WU, et al. The EANM practice guidelines for bone scintigraphy. Eur J Nucl Med Mol Imaging. 2016;43(9):1723-1738. doi:10.1007/s00259-016-3415-4. doi.org
- Dickson JC, Armstrong IS, Gabiña PM, et al. EANM practice guideline for quantitative SPECT-CT. Eur J Nucl Med Mol Imaging. 2023;50(4):980-995. doi:10.1007/s00259-022-06028-9. doi.org
- 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. doi.org
- Bailey DL, Willowson KP. Quantitative SPECT/CT: SPECT joins PET as a quantitative imaging modality. Eur J Nucl Med Mol Imaging. 2014;41(Suppl 1):S17-S25. doi:10.1007/s00259-013-2542-4. 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
- Zeintl J, Vija AH, Yahil A, Hornegger J, Kuwert T. Quantitative accuracy of clinical 99mTc SPECT/CT using ordered-subset expectation maximization with 3-dimensional resolution recovery, attenuation, and scatter correction. J Nucl Med. 2010;51(6):921-928. doi:10.2967/jnumed.109.071571. doi.org
- Ritt P, Vija H, Hornegger J, Kuwert T. Absolute quantification in SPECT. Eur J Nucl Med Mol Imaging. 2011;38(Suppl 1):S69-S77. doi:10.1007/s00259-011-1770-8. doi.org
- Yamane T, Fukushima K, Shirotake S, et al. Test-retest repeatability of quantitative bone SPECT/CT. Ann Nucl Med. 2021;35(3):338-346. doi:10.1007/s12149-020-01568-2. doi.org
- Kaneta T, Ogawa M, Daisaki H, et al. SUV measurement of normal vertebrae using SPECT/CT with Tc-99m methylene diphosphonate. Am J Nucl Med Mol Imaging. 2016;6(5):262-268. PMID:27766184. PubMed
- Ichikawa H, Miyaji N, Onoguchi M, et al. Feasibility of ultra-high-speed acquisition in xSPECT bone algorithm: a phantom study with advanced bone SPECT-specific phantom. Ann Nucl Med. 2022;36(2):183-190. doi:10.1007/s12149-021-01689-2. doi.org
- Arvola S, Jambor I, Kuisma A, et al. Comparison of standardized uptake values between 99mTc-HDP SPECT/CT and 18F-NaF PET/CT in bone metastases of breast and prostate cancer. EJNMMI Res. 2019;9(1):6. doi:10.1186/s13550-019-0475-z. doi.org
- Tabotta F, Jreige M, Schaefer N, Becce F, Prior JO, Nicod Lalonde M. Quantitative bone SPECT/CT: high specificity for identification of prostate cancer bone metastases. BMC Musculoskelet Disord. 2019;20(1):619. doi:10.1186/s12891-019-3001-6. doi.org
- Zhang Y, Li B, Yu H, et al. The value of skeletal standardized uptake values obtained by quantitative single-photon emission computed tomography-computed tomography in differential diagnosis of bone metastases. Nucl Med Commun. 2021;42(1):63-67. doi:10.1097/MNM.0000000000001311. doi.org
- Guo J, Chen C, Xiong L, et al. Analysis of the application value of SUV in quantitative SPECT/CT for diagnosing benign and malignant bone lesions. Hell J Nucl Med. 2024;27(3):198-205. doi:10.1967/s002449912754. doi.org
Related Articles
-
Gamma Camera Uniformity QC: Floods and Limits
Gamma camera flood-field uniformity QC explained: NEMA integral and differential uniformity, intrinsic vs extrinsic floods, count statistics, and limits.
-
Gamma Camera Detector Physics and Performance
How a gamma camera actually works: the NaI(Tl) crystal, PMT array, Anger positioning, pulse-height analysis, and the physics behind intrinsic resolution.
-
SPECT Jaszczak Phantom QC: Resolution & Contrast
The Jaszczak SPECT phantom checks reconstructed resolution, cold-sphere contrast, and tomographic uniformity. Learn how to fill, acquire, analyze, and pass it.