Cardiac SPECT MPI Physics and Quality Control

Cardiac SPECT myocardial perfusion imaging is the most widely performed nuclear cardiology procedure in the United States, and its diagnostic accuracy depends on the rigor of every step in the imaging chain—from radiopharmaceutical selection and stress protocol through acquisition, reconstruction, correction, and image interpretation. A board-certified medical physicist supporting a cardiac SPECT program must understand not only how to test a gamma camera but also how the camera interacts with the cardiac-specific workflow: the tracer pharmacokinetics that determine timing, the collimator and orbit choices that set resolution, the ECG gating that enables functional analysis, and the artifact patterns that can mislead interpretation.

This post covers the physics, quality-control requirements, and practical optimization strategies for cardiac SPECT MPI, providing a framework that aligns with ASNC and SNMMI imaging guidelines, NEMA NU-1 performance standards, and AAPM and IAEA QC guidance. DRPS provides these services as part of its PET/CT and nuclear medicine physics and accreditation support programs across Florida, Maryland, Virginia, Washington DC, California, and Nevada.

Introduction

Cardiac SPECT myocardial perfusion imaging evaluates regional myocardial blood flow at stress and rest, providing diagnostic and prognostic information in patients with known or suspected coronary artery disease. The examination uses a gamma camera equipped with low-energy high-resolution (LEHR) collimators to acquire multiple planar projections as the detector rotates around the patient; those projections are then reconstructed tomographically into three-dimensional perfusion maps of the left ventricular myocardium.

What makes cardiac SPECT MPI physically demanding is that the myocardium is a small, moving structure surrounded by high-activity background organs—the liver, spleen, bowel, and stomach—that compete with the heart for detected counts. Small changes in photon attenuation, patient or cardiac motion, or reconstruction parameters produce characteristic artifacts that can be misread as perfusion defects or that can obscure real ones. ^{1, 2}

The governing clinical guidance is the ASNC Imaging Guidelines for Nuclear Cardiology Procedures and the SNMMI Procedure Standard for Myocardial Perfusion Imaging, which define minimum acquisition parameters, quality control expectations, and interpretation criteria. The instrument performance framework is NEMA NU-1, and the QC methodology is elaborated in AAPM Report No. 22 and IAEA Human Health Series No. 6. ^{3, 4, 5, 6}

Topic Explanation

What is cardiac SPECT MPI and why does physics matter?

Cardiac SPECT MPI uses a radiolabeled tracer that distributes in proportion to regional myocardial blood flow, so regions with reduced perfusion (from coronary stenosis or prior infarction) show as areas of decreased uptake in the reconstructed images. Stress images acquired during or immediately after exercise or pharmacologic vasodilation reveal flow-limiting stenoses that may not be apparent at rest. The comparison of stress and rest images distinguishes ischemic (reversible) from infarcted (fixed) perfusion defects.

The imaging chain has several steps, each introducing physics-based performance limits:

Radiopharmaceutical: determines photon energy, administered activity, dosimetry, background distribution, and timing requirements.
Gamma camera and collimator: set the detection efficiency, spatial resolution, and energy discrimination.
Acquisition orbit and protocol: determine angular sampling, time per projection, and motion sensitivity.
Reconstruction: converts projections into tomographic slices and may incorporate corrections for resolution, attenuation, and scatter.
Corrections: attenuation and scatter correction reduce systematic artifacts that can mimic or mask disease.
ECG gating: enables functional analysis and helps distinguish true perfusion defects from motion and attenuation artifacts.
Quality control: confirms at each step that the system is performing within tolerance.

Understanding this chain is necessary for diagnosing equipment failures, optimizing protocols, and advising clinical teams on artifact recognition and interpretation limitations.

Key Technical Principles

Radiopharmaceuticals: Tc-99m agents and Tl-201

The choice of radiopharmaceutical establishes the photon energy, administered activity, patient dosimetry, organ background, and timing constraints of the entire examination.

Technetium-99m–labeled agents — Tc-99m sestamibi (MIBI) and Tc-99m tetrofosmin — are the clinical workhorses of cardiac SPECT MPI. Both are lipophilic monovalent cations that diffuse passively across sarcolemmal and mitochondrial membranes and are sequestered in viable myocytes in proportion to regional blood flow, retained by the large negative mitochondrial transmembrane potential, with minimal redistribution after the initial extraction. First-pass myocardial extraction fraction is moderate (roughly 60% for sestamibi and 54% for tetrofosmin), so net uptake underestimates true flow at high flow rates — a "roll-off" that limits sensitivity for balanced multivessel disease but is acceptable for routine relative-perfusion imaging. The 140.5 keV gamma emission of Tc-99m (≈89% abundance) closely matches the peak intrinsic detection efficiency of a ~9.5 mm NaI(Tl) crystal and the design energy of LEHR collimators, yielding higher count rates and better spatial resolution than Tl-201. The 6.0-hour physical half-life permits administered activities an order of magnitude higher than Tl-201 at comparable effective dose, supporting same-day rest-stress protocols and high-count-density ECG-gated acquisition. According to PubMed, Tc-99m-labeled agents offer improved spatial resolution, shorter half-life, and higher photon energy than Tl-201, enabling faster imaging and better image quality while preserving the ability to assess myocardial viability when combined with gated SPECT and nitrate augmentation (Knott et al., 2025; Caner & Beller, 1998). ^{7, 8}

Thallium-201 was the original MPI tracer and remains in selective use for stress-redistribution-reinjection viability protocols. Tl-201 is a potassium analog actively transported into myocytes through the Na⁺/K⁺-ATPase pump, with a high first-pass extraction fraction (~85%). Unlike the Tc-99m agents, it redistributes over hours from high-flow to low-flow but still-viable regions, so a single stress injection followed by a 3–4 hour delayed image can separate ischemia (which fills in) from scar (which does not). Tl-201 decays by electron capture; its imageable signal is dominated by the 68–80 keV mercury characteristic X-rays (≈94% combined abundance), with minor gamma lines at 135 keV and 167 keV. The low photon energy increases tissue attenuation and Compton scatter and degrades intrinsic spatial resolution relative to Tc-99m. The 73.0-hour physical half-life is the principal reason Tl-201 carries a substantially higher patient effective dose — on the order of 0.11–0.14 mSv/MBq versus roughly 0.0079–0.009 mSv/MBq for the Tc-99m agents per ICRP Publication 80 — which caps administered activity at a few millicuries. ^{7, 8, 15}

The table below summarizes the key imaging characteristics of the three agents. Administered-activity figures are representative of ASNC-protocol practice; ASNC and SNMMI explicitly favor weight-based dosing and the lowest activity consistent with diagnostic image quality, so a given program's values should be set against its own protocol, camera sensitivity, and dose-calibrator records.

Property	Tc-99m sestamibi	Tc-99m tetrofosmin	Tl-201
Principal imageable photons	140.5 keV γ (≈89%)	140.5 keV γ (≈89%)	68–80 keV Hg X-rays (≈94%); 167 keV γ (≈10%)
Physical half-life	6.0 hours	6.0 hours	73.0 hours
Decay mode	Isomeric transition	Isomeric transition	Electron capture
Myocardial uptake mechanism	Passive diffusion, mitochondrial trapping	Passive diffusion, mitochondrial trapping	Na⁺/K⁺-ATPase active transport
First-pass extraction fraction	~60%	~54%	~85%
Redistribution	Minimal	Minimal	Yes (1–4 h)
Same-day protocol	Yes (rest-stress or stress-rest)	Yes (rest-stress or stress-rest)	Stress-redistribution (3–4 h) ± reinjection
Representative stress activity	~740–1110 MBq (20–30 mCi)	~740–1110 MBq (20–30 mCi)	~74–148 MBq (2–4 mCi)
Effective dose coefficient (ICRP 80)	~0.0079 mSv/MBq	~0.0076 mSv/MBq	~0.14 mSv/MBq (rest)
Relative image quality	Good–excellent	Good–excellent	Moderate (low energy, scatter)
Primary use	Stress/rest MPI, viability (gated)	Stress/rest MPI	Viability, stress-redistribution

Stress protocols: 1-day, 2-day, and Tl-201 redistribution

The timing and dose ratio between rest and stress acquisitions are determined by the agent's redistribution behavior and by the need to minimize residual background from one acquisition interfering with the other.

For Tc-99m agents, the protocol design follows directly from two physics constraints: the 6-hour half-life is too long to let the first injection's counts decay away before the second, so the two acquisitions must be separated by activity ratio or by a full day; and the tracers clear through the hepatobiliary system, so imaging must wait for subdiaphragmatic background to fall. The main designs are:

1-day rest-stress protocol: a lower rest dose (typically ~296–444 MBq, 8–12 mCi) is injected first and imaged; a higher stress dose at a ≥3:1 activity ratio (~925–1110 MBq, 25–30 mCi) is injected at peak stress and imaged second. The ratio ensures the stress image is dominated by fresh stress counts rather than residual rest activity. Because residual rest activity still adds a low background, the stress study should be read with that in mind. Per the sestamibi and tetrofosmin package inserts and ASNC guidance, imaging begins roughly 15–60 minutes after injection — closer to 45–60 minutes for sestamibi at rest (slower hepatic clearance) and as early as 15–30 minutes for tetrofosmin, with stress images obtainable sooner than rest images because exercise diverts blood flow from the splanchnic bed and accelerates relative cardiac-to-liver contrast.
2-day protocol: equal higher doses (~925–1110 MBq, 25–30 mCi) on separate days eliminate cross-contamination entirely and provide the best count statistics; it is preferred for larger patients (body mass index roughly >30 kg/m²) where attenuation demands maximum counts, at the cost of a second visit.
1-day stress-first (stress-rest) protocol: a stress-only "gatekeeper" acquisition is read first; if perfusion and gated wall motion are normal, the rest injection is omitted, sparing the patient a second dose and shortening the study.

For Tl-201, the classic stress-redistribution protocol injects ~74–111 MBq (2–3 mCi) at peak stress, acquires stress images promptly (within ~10–15 minutes, before significant redistribution begins), and re-images at 3–4 hours to capture redistribution. A rest reinjection of ~37 MBq (1 mCi) before the delayed acquisition improves viability sensitivity by re-delivering tracer to severely hypoperfused but viable segments that redistribute slowly.

Gamma-camera and SPECT acquisition parameters

Cardiac SPECT MPI acquisition quality depends on collimator selection, orbit geometry, angular sampling, and time per projection—parameters that collectively determine spatial resolution and count density in the reconstructed image.

Collimator: Low-energy high-resolution (LEHR) parallel-hole collimators are standard for Tc-99m cardiac SPECT, giving system spatial resolution on the order of 7–10 mm FWHM at the ~15–20 cm heart-to-collimator distance typical of a body-contour orbit. For Tl-201, low-energy all-purpose (LEAP/LEGP) collimators are sometimes substituted to recover sensitivity at the lower photon energy. High-sensitivity collimators trade resolution for count rate and are not recommended for routine cardiac SPECT.
Orbit: Most conventional cardiac SPECT uses a 180° orbit from 45° right anterior oblique (RAO) to 45° left posterior oblique (LPO). Because the heart sits in the left anterior hemithorax, this arc keeps the detector closest to the heart and maximizes counts and resolution; the trade-off is mild non-uniformity of resolution between opposing walls compared with a 360° orbit. ASNC guidelines accept 180° acquisition as the clinical standard for SPECT MPI, and a body-contour (noncircular) orbit is preferred over a fixed-radius circular orbit to minimize detector-to-patient distance throughout the arc.
Angular sampling: 60–64 projections over 180° (≈3° steps) is typical. Adequate angular sampling avoids streak/aliasing artifacts; finer sampling improves the reconstruction at the cost of acquisition time.
Time per projection: total counts and per-projection time are balanced against patient-motion risk, targeting on the order of a few million total counts in the cardiac field of view. A 15–20 minute total acquisition (≈20–25 s/projection) is common for standard dual-head Tc-99m protocols. Dedicated cadmium-zinc-telluride (CZT) cardiac cameras — multiple pinhole or stationary detector banks focused on the heart — achieve 5- to 8-fold higher count sensitivity, enabling acquisitions of roughly 4–8 minutes (or proportional dose reduction) with equivalent or superior count statistics and improved energy resolution.
Matrix and pixel size: a 64 × 64 matrix with ~6.4 mm pixels is standard for conventional SPECT; the pixel size should be roughly one-third to one-half the system FWHM to avoid undersampling resolution-recovery gains.

Gamma-camera and SPECT QC for MPI

The gamma-camera QC program for cardiac SPECT MPI builds on the standard NEMA NU-1–based QC schedule but emphasizes the tests most critical to cardiac image quality.

For a full treatment of the gamma-camera QC test set (uniformity, energy peaking, center of rotation, spatial resolution, sensitivity, multi-detector registration), see SPECT/CT Quality Control: Uniformity, Center of Rotation, and System Performance. Here we highlight the cardiac-specific considerations.

QC Test	Frequency	Cardiac-specific relevance
Energy peaking	Daily, each isotope in use	Tc-99m (140 keV) and Tl-201 require separate photopeak settings. A ~15–20% symmetric window centered on 140 keV is standard for Tc-99m; Tl-201 is imaged on a ~20–30% window over the 68–80 keV X-ray peak (often with a second window on the 167 keV line). Confirm window widths against your camera's clinical protocol.
Field uniformity (extrinsic flood)	Daily	Visual/quantitative flood; non-uniformities of more than a few percent back-project into concentric ring artifacts in transaxial cardiac slices
High-count flood uniformity	Weekly	Formal NEMA NU-1 method (≥30 million counts); UFOV/CFOV integral and differential uniformity within vendor specification, the floor for tomographic uniformity
Center of rotation (COR)	Weekly or per protocol	A COR offset of even ~0.5 pixel blurs the point spread and produces "tuning-fork" / doughnut artifacts on short-axis slices; re-verify after any detector or gantry maintenance
Spatial resolution and linearity	Weekly / monthly	Four-quadrant bar phantom; visual check of resolving power and spatial linearity; guards against PMT gain drift
SPECT phantom (Jaszczak-type)	Quarterly / annual	Tomographic uniformity, cold-rod resolution, sphere contrast; ACR and IAC accreditation require phantom images scored within tolerance
Dose calibrator QC	Constancy each day of use; linearity quarterly; accuracy and geometry at installation and after repair (annual accuracy verification per common practice)	Every administered activity is measured here; an inaccurate calibrator corrupts both the 3:1 stress:rest ratio and patient dosimetry. NRC removed the prescriptive instrument-QC schedule from §35.60; current 10 CFR 35.60–35.61 require a calibrated/checked instrument with QC performed per manufacturer or nationally recognized standards (e.g., NIST traceability, ANSI N42.13).
System sensitivity constancy	Annual (trend)	Stable sensitivity (cpm/MBq) confirms detector and electronics health; required for quantitative and serial MPI programs

Reconstruction: FBP versus iterative OSEM with resolution recovery

Filtered back-projection (FBP) was the historical reconstruction standard because of its computational speed and analytic predictability, but ordered-subset expectation maximization (OSEM) with resolution recovery has become the clinical standard for cardiac SPECT MPI because it provides better spatial resolution and contrast at typical count levels. ⁹

FBP applies a ramp filter (which compensates for the 1/r oversampling of central frequencies in back-projection) plus a low-pass apodization window — Butterworth or Hann — directly to the projection data before back-projecting across the image matrix. It is fast, linear, and analytically predictable, but the ramp filter amplifies high-frequency Poisson noise in low-count studies, forcing aggressive smoothing that degrades spatial resolution. FBP also cannot natively incorporate attenuation, scatter, or resolution physics, and it produces characteristic streak artifacts radiating from high-activity foci such as the liver or bowel.
OSEM is an iterative maximum-likelihood algorithm. At each step the current image estimate is forward-projected, compared to the measured projections, and updated by the multiplicative expectation-maximization (MLEM) correction, accelerated by dividing the projections into ordered subsets (OS) so the image is updated several times per full pass through the data. The MLEM update for voxel $j$ is:

λ_{j}^{(n + 1)} = \frac{λ _{j}^{(n)}}{\sum _{i} a _{ij}} i \sum a_{ij} \frac{p _{i}}{\sum _{k} a _{ik} λ _{k}^{(n)}}

where $λ_{j}^{(n)}$ is the activity estimate in voxel $j$ at iteration $n$ , $p_{i}$ is the measured counts in projection bin $i$ , and $a_{ij}$ is the system matrix element giving the probability that a decay in voxel $j$ is detected in bin $i$ . Because $a_{ij}$ embeds the imaging physics, attenuation, scatter, and collimator response can all be folded directly into the reconstruction — the key advantage over FBP. The number of effective iterations is (iterations × subsets); roughly 8–12 subsets with a small iteration count is typical.

Resolution recovery (RR), also called point-spread-function (PSF) or collimator-detector response (CDR) correction, encodes the distance-dependent detector blur into the system matrix $a_{ij}$ , so the forward and back-projection steps model — and partially undo — the geometric spreading of counts with source-to-collimator distance. This sharpens resolution and recovers contrast in small structures such as the thin myocardial wall, partially offsetting partial-volume losses. An observer study of reduced-count cardiac SPECT found that reconstructions using resolution recovery with only post-reconstruction smoothing had lower defect-detection AUC than strategies that also applied attenuation and scatter correction — and that diagnostic performance could be preserved at substantially reduced counts when full physics modeling was combined with appropriate denoising (Pretorius et al., 2023). ⁹

The trade-off is that OSEM has no natural stopping point: high-frequency noise and Gibbs-type edge overshoot grow with effective iterations as the algorithm converges toward the noisy data. Clinical cardiac SPECT therefore uses validated iteration/subset combinations (for example, ≈2–4 iterations × 10 subsets) with a post-reconstruction Butterworth filter to balance resolution against noise. Vendor-supplied protocols should be validated before clinical use and must not be altered without re-validation against phantom data, because changing iterations, subsets, or filter cutoff silently shifts both quantitative perfusion scores and apparent defect size.

Poisson count noise and clinical implications

A fundamental limit on SPECT image quality is Poisson counting noise. For a voxel that accumulates $N$ counts, the relative standard deviation of the count estimate is:

σ_{rel} = \frac{1}{N}

because the variance of a Poisson process equals its mean, $σ_{N} = N$ . Doubling the administered activity or scan time doubles $N$ and reduces relative noise by $1/ 2 \approx 0.71$ (a ~29% improvement); halving the dose increases relative noise by $2$ (a ~41% degradation). For cardiac SPECT, reconstructed myocardial-wall voxels typically hold a few hundred to a few thousand counts; a segment with only 200 counts carries $1/ 200 \approx 7%$ relative noise before any smoothing, setting a practical floor on contrast-to-noise ratio. This is why dose-reduction strategies must be validated carefully against defect-detection performance, and why high-sensitivity CZT systems can lower the dose substantially — the Pretorius observer study above supports preserved diagnostic accuracy at markedly reduced counts when attenuation, scatter, and resolution corrections are applied — rather than trading counts one-for-one against image quality. ⁹

Attenuation correction and scatter correction

Photon attenuation is the dominant source of systematic artifact in cardiac SPECT MPI. As 140 keV photons (or 68–80 keV for Tl-201) travel from the myocardium to the detector through soft tissue, primary counts fall off exponentially as $I = I_{0} e^{- μx}$ , where the linear attenuation coefficient $μ$ for soft tissue is roughly 0.15 cm⁻¹ at 140 keV and about 0.19 cm⁻¹ at the Tl-201 X-ray energies. Over ~10 cm of overlying tissue this attenuates a 140 keV photon beam by more than threefold; the lower-energy Tl-201 photons are attenuated even more, which is why attenuation artifacts are more pronounced with thallium. Uncorrected, this depth-dependent loss underestimates counts in deeper myocardial walls and creates apparent perfusion defects that track patient body habitus rather than coronary anatomy.

The most important clinically recognizable attenuation patterns are:

Inferior wall attenuation (diaphragmatic): more common in men, produces a fixed inferior wall defect that mimics or masks true inferior ischemia.
Anterior/anterolateral wall attenuation (breast tissue): more common in women, produces anterior or anterolateral defects. With upright CZT imaging, pendant breast tissue can instead produce an inferior-wall attenuation artifact — a pattern distinct from the anterior defect seen with conventional supine positioning. ¹⁰
Lateral wall attenuation: can occur in obese patients.

CT-based attenuation correction (CTAC) uses a low-dose CT acquired immediately before or after the SPECT scan on a hybrid SPECT/CT system. The CT Hounsfield-unit map is converted, via a bilinear calibration, into a map of linear attenuation coefficients and then scaled from the effective CT beam energy (~70 keV) to the SPECT photon energy (140 keV for Tc-99m) before being incorporated into the iterative reconstruction system matrix. CTAC is the most accurate attenuation-correction method clinically available, but it introduces a new failure mode: misregistration between the CT map and the SPECT emission data, usually from differing respiratory states or patient motion between the two acquisitions. Even a 1-cm shift can place the CT lung over the SPECT myocardium and create an artifactual anterior or lateral "defect," or, conversely, place liver over the inferior wall and overcorrect it. Every CTAC study should therefore be reviewed with the emission and transmission images fused before the attenuation-corrected images are trusted.

Line-source / scanning line-source correction uses external sealed radioactive transmission sources (e.g., Gd-153, ~100 keV) instead of CT to generate the attenuation map. It avoids the CT dose and the respiratory-timing mismatch but yields a lower-resolution, noisier map and is now largely superseded by CTAC on hybrid systems.

Scatter correction removes counts from photons that Compton-scattered before detection — these are accepted into the photopeak window because the camera's finite energy resolution (~9–10% FWHM for NaI(Tl) at 140 keV) cannot reject all degraded photons, and they reduce image contrast and bias attenuation-corrected quantification. The dual- and triple-energy-window (DEW/TEW) methods are the common clinical approaches: counts in narrow windows abutting the photopeak estimate the in-peak scatter, which is then subtracted (or, better, modeled within OSEM). In that observer study, reconstructions using resolution recovery alone showed lower defect-detection AUC than those that added attenuation and scatter correction — evidence that modeling both is worthwhile, particularly in reduced-count imaging (Pretorius et al., 2023). ⁹

ECG-gated SPECT and LVEF

ECG-gated SPECT acquires projection data synchronized to the cardiac cycle via R-wave triggering, enabling reconstruction of the heart at multiple phases and direct measurement of left ventricular (LV) systolic function. This adds significant clinical value to the perfusion study because LV ejection fraction (LVEF), end-diastolic volume (EDV), end-systolic volume (ESV), and regional wall motion are all derived from the same acquisition. ¹¹

Standard gating divides each R–R interval into 8 or 16 frames, with counts from many hundreds of cardiac cycles summed into the corresponding frames to build adequate statistics per gate. The gated projections are reconstructed into a cine series, and surface-fitting algorithms (such as the Cedars-Sinai QGS algorithm) automatically detect the endocardial and epicardial surfaces at each gate. LVEF is computed from the extracted volumes as:

LVEF = \frac{EDV - ESV}{EDV} \times 100%

where EDV is the end-diastolic volume (maximum LV cavity volume, near the first gate) and ESV is the end-systolic volume (minimum cavity volume, near mid–R–R). According to PubMed, low-dose Tc-99m-tetrofosmin gated SPECT on a CZT camera correlated excellently with cardiac MRI for LV volumes and ejection fraction but significantly underestimated absolute volumes; 16-frame reconstruction gave a higher LVEF correlation (r = 0.91) and a mean EF statistically indistinguishable from MRI, versus 8-frame (r = 0.84) which significantly underestimated EF (Giorgetti et al., 2013). The trade-off is that 16-gate acquisition splits the same counts into twice as many frames, so it requires adequate count density to avoid noise-driven contour errors. ¹²

Clinical value of gated SPECT includes:

Artifact differentiation: a fixed perfusion defect with matching akinesis/hypokinesis and loss of systolic wall thickening is more likely true scar; a fixed apparent defect with preserved wall motion and thickening is more likely an attenuation artifact, because viable myocardium thickens even where overlying tissue suppresses its counts. The brightening-with-thickening cue (a partial-volume effect) is one of the most powerful single-acquisition discriminators in clinical practice. ^{8, 11}
Incremental prognostic information: LVEF and LV volumes add independent prognostic value to perfusion data.
Functional assessment after revascularization: gated SPECT is used to monitor LV function recovery.

Gating quality should be verified: irregular heart rhythms (atrial fibrillation, frequent ectopy) degrade gated image quality by causing non-uniform R–R interval binning. Beat acceptance windows (typically ±20% of mean R–R) reject irregular beats; if too many beats are rejected, count statistics in the gated images are reduced.

Clinical Impact

The practical consequence of physics failures in cardiac SPECT MPI is diagnostic error. Each type of artifact maps to a recognizable clinical risk.

Attenuation artifacts masquerading as perfusion defects are the most common cause of false-positive cardiac SPECT studies. An uncorrected inferior wall defect in a male patient without clinical risk factors is classic; without CTAC or gated SPECT wall motion data, the interpreter must use clinical judgment, often leading to unnecessary downstream testing.
Breast attenuation producing an anterior or anterolateral apparent defect in women is a persistent challenge that has driven interest in CTAC and upright-positioning systems. ¹⁰
Patient motion during acquisition smears projection data across angles, blurring myocardial walls and potentially creating or obscuring defects. Sinogram review and motion correction are important QA steps.
Upward creep after treadmill exercise causes the diaphragm and heart to move superiorly as breathing normalizes post-stress; projections acquired during this period are motion-contaminated, producing an apparent inferior wall defect. The standard mitigation is a brief delay (typically 15–30 minutes) between stress peak and stress acquisition, or use of pharmacologic stress.
High subdiaphragmatic background from liver or bowel activity in sestamibi studies can spill into the inferior wall, creating an apparent perfusion defect or obscuring a true one; acquisition timing (waiting for adequate hepatic clearance) and patient positioning are key mitigations.
Reconstruction artifacts: too many OSEM iterations can create edge overshoot (Gibbs artifact) at the myocardial boundary; too few produce blurring. COR errors in the underlying gamma-camera produce ring or blur artifacts in myocardial slices.

The medical physicist's role is to ensure the QC program catches equipment failures before they cause diagnostic errors, to validate reconstruction protocols, to advise on artifact recognition, and to support protocol optimization for challenging patient populations.

Practical Optimization Tips

A well-run cardiac SPECT MPI program manages five areas: dose protocol, acquisition, reconstruction validation, QC schedule, and artifact awareness.

Dose protocol and timing

Follow the current ASNC imaging guidelines for administered-activity selection; use weight-based dosing and the minimum activity consistent with diagnostic image quality for the patient's body habitus, camera sensitivity, and protocol. CZT cameras and resolution-recovery reconstruction are the primary levers for dose reduction without losing diagnostic accuracy.
For sestamibi: allow adequate hepatobiliary clearance (typically ~45–60 minutes post-injection at rest); a light fatty meal or water loading can accelerate clearance of subdiaphragmatic activity. For tetrofosmin, clearance is faster, so ~15–30 minutes is often acceptable.
Verify dose-calibrator accuracy and constancy on each day of use; calibrator error directly determines whether the intended ≥3:1 stress:rest activity ratio is actually delivered in 1-day protocols, and it propagates into patient dosimetry.

Acquisition

Position the patient carefully: supine acquisition is standard, but supplemental prone acquisition can help distinguish inferior attenuation from true disease by changing the attenuation geometry.
Use a body-contour (circular or elliptical step-and-shoot, or continuous) orbit to minimize detector-to-heart distance across the 180° arc, maximizing geometric resolution.
Review raw projection data (sinogram) before reconstruction: patient motion, cardiac motion, upward creep, and high subdiaphragmatic activity are all visible in the sinogram and must be assessed before interpreting the reconstructed images.

Reconstruction validation

Validate vendor-supplied OSEM+RR protocols with a SPECT performance phantom (Jaszczak-type) before clinical use, and after any software update.
Confirm that attenuation correction and scatter correction are applied consistently and that the CT attenuation map is checked for misregistration artifacts on each study.
Use 8-gate or 16-gate ECG gating for every clinically meaningful Tc-99m study; verify gating statistics and beat acceptance fraction.

QC schedule and trending

Perform daily energy peaking on every isotope in use that day; the Tl-201 window requires separate verification from Tc-99m.
Trend all QC metrics (integral uniformity, COR offset, energy resolution, sensitivity) over time; slow drift identifies a component approaching failure.
Document and date-stamp every QC result; accreditation programs require organized QC logs, and trending requires longitudinal data.

Common pitfalls to avoid

Never skip gating on a Tc-99m stress study; it is the single most powerful artifact-mitigation and incremental-value tool available without additional dose.
Never interpret a cardiac SPECT study without reviewing the raw projection data first; artifacts visible in the sinogram cannot always be detected after reconstruction.
Avoid assuming attenuation correction eliminates all attenuation artifacts; CT misregistration can introduce its own artifacts and should be checked on every CTAC study.
Do not change reconstruction parameters (iterations, subsets, filter settings) without re-validating the protocol against phantom data and confirming consistency with prior studies in the trend database.

Regulatory Considerations

Cardiac SPECT MPI programs operate under two regulatory regimes: the NRC (or Agreement State) for the radioactive byproduct material used as the radiopharmaceutical, and the FDA and state radiation-control authority for the gamma camera as a radiation-producing device.

Key regulatory frameworks include:

10 CFR Part 35 (Medical Use of Byproduct Material): governs the authorized use of Tc-99m and Tl-201 in nuclear medicine, instrument calibration, the authorized-user/RSO structure, medical-event (formerly "misadministration") reporting, and radiation safety program obligations for facilities holding an NRC or Agreement State radioactive material license. ¹³
10 CFR Part 20: sets occupational dose limits, public dose limits, and ALARA requirements applicable to nuclear medicine staff handling and administering radiopharmaceuticals. ¹⁴
10 CFR 35.60–35.63: §35.60 requires that the instrument used to measure activity (the dose calibrator) be calibrated and that activity-measurement QC be performed in accordance with nationally recognized standards or the manufacturer's instructions — the NRC dropped the older prescriptive "daily constancy / quarterly linearity / annual accuracy" schedule from the regulation, though that schedule remains common best practice and is often retained in license conditions. §35.63 requires that the administered activity of every dosage be determined (measured or decay-calculated) before administration. Together these anchor the administered-activity accuracy that underpins protocol dose ratios and patient dosimetry.
ASNC and SNMMI guidelines: while not regulatory mandates, ASNC Imaging Guidelines for Nuclear Cardiology Procedures and SNMMI Procedure Standards define the standard of care for cardiac SPECT MPI and are referenced in accreditation and Joint Commission reviews. ^{3, 4}
ACR Nuclear Medicine and PET Accreditation Program: requires documented QC, qualified medical physicist oversight (including annual camera surveys), and phantom imaging within tolerance for accredited facilities. Accreditation often conditions Medicare reimbursement eligibility.

Jurisdiction: of the states DRPS serves, Florida, Maryland, Virginia, California, and Nevada are NRC Agreement States that administer radioactive material licensing under their own radiation-control regulations. Washington, DC is regulated directly by the NRC. Each Agreement State enforces rules equivalent to 10 CFR Parts 20 and 35, but facilities must confirm the applicable state rules before assuming federal CFR text applies verbatim. For related regulatory context, see Dose Calibrator Quality Control.

Frequently Asked Questions (FAQs)

What radiopharmaceuticals are used for cardiac SPECT MPI?

The two most common agents are Tc-99m sestamibi (MIBI) and Tc-99m tetrofosmin, both technetium-labeled agents imaged on the 140 keV gamma photopeak. Thallium-201 (Tl-201), imaged on its 68–80 keV mercury X-ray peak, was the original MPI tracer and is still used selectively for viability and stress-redistribution protocols, though its higher patient dose and lower image quality have made the Tc-99m agents the clinical standard.

What is the difference between a 1-day and a 2-day cardiac SPECT protocol?

A 1-day protocol injects rest and stress doses on the same day, with the rest dose given first (lower activity) and the stress dose given second (higher activity, typically 3:1 ratio) to minimize residual background from the rest study. A 2-day protocol uses equal higher doses on separate days, providing better count statistics and image quality at the cost of patient convenience.

Why does cardiac SPECT need attenuation correction?

Photon attenuation from overlying tissue—especially the diaphragm, breast, and myocardium itself—reduces counts in the inferior and anterior walls, creating apparent perfusion defects that can mimic coronary artery disease. CT-based attenuation correction (or line-source correction) measures the actual attenuation map for each patient and corrects projection data before or during reconstruction, reducing these false-positive artifacts.

What is ECG-gated SPECT and how is LVEF calculated?

ECG-gated SPECT acquires projection data binned into 8 or 16 gates synchronized to the R–R interval of the ECG, allowing reconstruction of the heart at different points in the cardiac cycle. Software algorithms detect the endo- and epicardial surfaces at end-diastole and end-systole to compute left ventricular volumes, from which LVEF = (EDV − ESV)/EDV × 100% is derived.

What QC tests are specific to cardiac SPECT MPI?

Beyond the standard gamma-camera QC program (daily energy peaking and uniformity, high-count flood, COR, bar phantom), cardiac SPECT MPI programs should include SPECT phantom imaging (Jaszczak-type) for tomographic uniformity and contrast, sensitivity constancy checks, and verification of the dose calibrator used to measure administered activity. ASNC and SNMMI guidelines specify minimum QC expectations for cardiac imaging programs.

What causes the upward creep artifact in cardiac SPECT?

Upward creep occurs because deep breathing immediately after stress moves the diaphragm and thus the heart upward relative to where it was during peak exercise. If the first projections are acquired before the patient's breathing pattern has normalized, the heart position shifts during the rotation, blurring the inferior wall and potentially mimicking an inferior perfusion defect.

How does OSEM with resolution recovery improve cardiac SPECT image quality?

Ordered-subset expectation maximization (OSEM) with resolution recovery (also called point-spread-function or PSF correction) models the collimator-detector response during reconstruction, sharpening spatial resolution and improving contrast compared to filtered back-projection. Combined with attenuation and scatter correction, OSEM with resolution recovery is the current clinical standard for cardiac SPECT MPI reconstruction.

Key Takeaways

Radiopharmaceutical selection determines the imaging chain. Tc-99m sestamibi and tetrofosmin (140 keV, minimal redistribution, same-day protocols) are the clinical standard; Tl-201 (68–80 keV X-rays, redistribution, higher dose) retains a role in viability and stress-redistribution studies.
LVEF from gated SPECT is essential, not optional. The equation $LVEF = (EDV - ESV) / EDV \times 100%$ derived from ECG-gated reconstruction adds independent diagnostic and prognostic value and is the most powerful single-acquisition tool for distinguishing attenuation artifacts from true perfusion defects.
Poisson counting statistics set the floor on image quality: $σ_{rel} = 1/ N$ . Dose, sensitivity, and scan time must be balanced to achieve adequate count density in myocardial wall voxels.
OSEM with resolution recovery plus attenuation and scatter correction is the reconstruction standard. FBP remains a fallback but produces inferior image quality at typical cardiac SPECT count levels.
Attenuation and scatter correction reduce but do not eliminate artifacts. CT misregistration, breast tissue positioning, and subdiaphragmatic activity create new artifact patterns that require systematic review of the raw data on every study.
QC must cover both the gamma camera and the dose calibrator. Dose calibrator accuracy directly controls the administered-activity accuracy, which underpins protocol dose ratios and patient dosimetry in every cardiac SPECT study.
Review raw projection data before interpreting reconstructed images. Patient motion, upward creep, and high background are visible in the sinogram and cannot always be detected after reconstruction.

Conclusion

Cardiac SPECT myocardial perfusion imaging is a technically demanding study that rewards systematic attention to physics at every step. The radiopharmaceutical sets the photon energy and timing constraints; the gamma camera and collimator set the detection geometry; the acquisition protocol determines count density and angular sampling; reconstruction with OSEM and resolution recovery sharpens spatial detail; attenuation and scatter corrections reduce systematic biases; and ECG gating adds functional information that transforms a perfusion map into a comprehensive cardiac evaluation. Quality control must span all of these layers—from the dose calibrator that measures every dose to the SPECT phantom that validates tomographic performance—and must be documented rigorously to support accreditation, clinical decision-making, and regulatory compliance.

The facilities that invest in this kind of systematic QC program—anchored in ASNC and SNMMI guidelines, built on NEMA NU-1 methodology, and validated with regular phantom imaging—protect their patients from diagnostic error and their programs from accreditation risk.

How DRPS Can Help

Diagnostic Radiation Physics Services provides board-certified medical physicist support for nuclear cardiology programs across Florida, Maryland, Virginia, Washington DC, California, and Nevada. Our services include PET/CT and nuclear medicine physics support encompassing gamma-camera acceptance testing, NEMA NU-1 annual surveys, SPECT phantom QC, reconstruction protocol validation, dose calibrator QC programs, and cardiac SPECT artifact consultation. We also provide accreditation support for ACR Nuclear Medicine and PET Accreditation, helping programs document QC, pass phantom requirements, and satisfy qualified-medical-physicist oversight expectations.

If you are commissioning a new cardiac SPECT program, troubleshooting image artifacts, validating an OSEM protocol update, or preparing for accreditation, our team can help. Contact DRPS or review our service locations to get started.

Related Resources

References

American Society of Nuclear Cardiology. ASNC Imaging Guidelines for Nuclear Cardiology Procedures: Myocardial Perfusion Imaging. Fairfax, VA: ASNC. asnc.org
Society of Nuclear Medicine and Molecular Imaging. SNMMI Procedure Standard for Myocardial Perfusion Imaging. Reston, VA: SNMMI. snmmi.org
National Electrical Manufacturers Association. NU 1: Performance Measurements of Gamma Cameras. Rosslyn, VA: NEMA. nema.org
American Association of Physicists in Medicine. Rotating Scintillation Camera SPECT Acceptance Testing and Quality Control (AAPM Report No. 22). College Park, MD: AAPM. aapm.org
American Association of Physicists in Medicine. Quantitation of SPECT Performance (AAPM Report No. 52). College Park, MD: AAPM. aapm.org
International Atomic Energy Agency. Quality Assurance for SPECT Systems. IAEA Human Health Series No. 6. Vienna: IAEA; 2009. iaea.org
Knott JD, Kronzer E, Anavekar N, Chareonthaitawee P, Askew JW. Radionuclide imaging techniques for assessing myocardial viability: Clinical applications, evidence, and future directions. Prog Cardiovasc Dis. 2025;93:51-59. doi:10.1016/j.pcad.2025.08.005. doi.org
Caner B, Beller GA. Are technetium-99m-labeled myocardial perfusion agents adequate for detection of myocardial viability? Clin Cardiol. 1998;21(4):235-42. doi:10.1002/clc.4960210402. doi.org
Pretorius PH, Liu J, Kalluri KS, et al. Observer studies of image quality of denoising reduced-count cardiac SPECT myocardial perfusion imaging by three-dimensional Gaussian post-reconstruction filtering and deep learning. J Nucl Cardiol. 2023;30(6):2427-2437. doi:10.1007/s12350-023-03295-3. doi.org
Waqar F, Athar MW, Dwivedi AK, et al. Visual patterns of breast attenuation artifacts in women and men with an upright and supine cadmium-zinc-telluride camera. J Nucl Cardiol. 2021;29(4):1976-1984. doi:10.1007/s12350-021-02632-8. doi.org
Noordzij W, Slart RH. Clinical value of quantitative measurements derived from GATED SPECT: motion and thickening, volumes and related LVEF. Q J Nucl Med Mol Imaging. 2016;62(3):321-324. doi:10.23736/S1824-4785.16.02868-X. doi.org
Giorgetti A, Masci PG, Marras G, et al. Gated SPECT evaluation of left ventricular function using a CZT camera and a fast low-dose clinical protocol: comparison to cardiac magnetic resonance imaging. Eur J Nucl Med Mol Imaging. 2013;40(12):1869-75. doi:10.1007/s00259-013-2505-9. doi.org
U.S. Nuclear Regulatory Commission. 10 CFR Part 35: Medical Use of Byproduct Material. ecfr.gov
U.S. Nuclear Regulatory Commission. 10 CFR Part 20: Standards for Protection Against Radiation. ecfr.gov
International Commission on Radiological Protection. ICRP Publication 80: Radiation Dose to Patients from Radiopharmaceuticals (Addendum 2 to ICRP Publication 53). Annals of the ICRP. 1998;28(3). icrp.org