Is phase analysis an approved standalone diagnostic test?

It is best understood as an adjunctive quantitative tool derived from a standard gated SPECT MPI study, not as a separate approved indication. The perfusion study is acquired and interpreted under the applicable imaging guidelines; phase analysis adds mechanical-timing information from the same dataset. Its outputs are vendor-specific and should be reported with the software and normal database identified.

Gated SPECT Phase Analysis for LV Dyssynchrony

Q: How is mechanical dyssynchrony different from electrical dyssynchrony?

Electrical dyssynchrony is inferred from the surface ECG, mainly QRS duration, and reflects the spread of electrical activation. Mechanical dyssynchrony is the actual dispersion in the timing of regional myocardial contraction. The two are related but not identical: patients with a wide QRS can contract relatively synchronously, and some patients with a narrow QRS show substantial mechanical dyssynchrony. Phase analysis measures the mechanical timing directly from the perfusion study.

Q: Are phase SD and bandwidth values comparable between software packages?

No. Normal limits and dyssynchrony cut-points are specific to the software, the normal database, and the acquisition protocol, and they are not interchangeable. A phase standard deviation reported by Emory Cardiac Toolbox cannot be compared directly with one reported by Quantitative Gated SPECT or another package, because reported normals can differ roughly two-fold between programs. Always interpret a value against the normal database of the same software used to generate it.

Q: Does 8-frame gating ruin phase analysis?

Not necessarily. Although an 8-frame gate samples the cardiac cycle only every 45 degrees, the multiharmonic Fourier fit interpolates the onset of contraction to well below one frame, and phantom work has shown temporal resolution on the order of a sixtieth of the cardiac cycle when counts are adequate. Adequate counts in each region matter more than frame number alone, but 16-frame gating is still generally preferred when count statistics allow.

Q: Can phase analysis predict who responds to cardiac resynchronization therapy?

Phase analysis has shown value in predicting response to cardiac resynchronization therapy in several single-center studies, with larger baseline dyssynchrony associated with a higher likelihood of response. However, the published cut-points are software-specific, derive from relatively small cohorts, and should be viewed as supportive rather than as a standalone selection criterion. Device-therapy decisions follow established clinical guidelines, with phase analysis as an adjunct.

Q: What counts as a significant phase SD or bandwidth?

There is no single universal threshold. In one Emory Cardiac Toolbox dataset, normal controls showed a phase SD near nine degrees and a bandwidth near twenty-nine degrees, while patients with severe LV dysfunction showed values several times higher. Because thresholds depend on software, sex, ejection fraction, and LV volume, a value should be judged against the matched normal database rather than a memorized number.

Every ECG-gated myocardial perfusion SPECT study already carries the information needed to measure left ventricular mechanical dyssynchrony — phase analysis simply extracts it. By fitting a Fourier function to the count-versus-time curve of each left-ventricular region, the software estimates when each region begins to contract, then displays those timings as a phase histogram. The width and disorder of that histogram, summarized by phase standard deviation, histogram bandwidth, and entropy, quantify how synchronously the ventricle contracts — at no additional acquisition time, radiation dose, or patient visit.¹⁸

A defensible phase-analysis report depends on understanding what the metrics mean, how they are computed, and a single fact that governs every number: phase-analysis normal limits and cut-points are software- and protocol-specific, and they are not interchangeable between packages.¹⁵

Introduction

Phase analysis of gated SPECT turns a routine perfusion study into a map of contraction timing, and it does so retrospectively from data you already have. Left ventricular mechanical dyssynchrony — a dispersion in when different regions of the ventricle begin to contract — is a marker of disease and, in selected patients, a potential predictor of response to cardiac resynchronization therapy (CRT).⁴⁹

The clinical motivation is concrete. Patient selection for CRT has historically relied heavily on QRS duration, an electrical surrogate, yet a substantial fraction of patients selected by electrical criteria do not improve. That mismatch has driven interest in measuring mechanical dyssynchrony directly.³⁴ Phase analysis is attractive because it is automated, reproducible, and free in the sense that no extra acquisition is required — the gated perfusion study acquired to assess perfusion and ejection fraction already contains the regional timing information.¹⁶

This article explains what gated SPECT phase analysis measures, the Fourier mathematics behind the phase histogram, the four histogram metrics and why their values are tied to a specific software package, the clinical evidence, practical reporting tips, and where the technique sits within current imaging guidance. Throughout, the guiding principle is that phase metrics are quantitative but context-dependent, and must be interpreted against the matched normal database.¹⁵¹⁶

Topic Explanation

What does gated SPECT measure that enables phase analysis?

A gated myocardial perfusion SPECT study divides the cardiac cycle, triggered by the ECG R wave, into a fixed number of frames — commonly 8 or 16. Reconstructing each frame produces a time series of three-dimensional perfusion images across the cycle. As the myocardium thickens during systole, the partial-volume effect makes counts in the wall rise and fall through the cycle. That count-versus-time behavior in each region is the raw signal phase analysis uses.¹⁸

Phase analysis samples the left ventricular wall into many regions and, for each region, measures how its maximal counts change through the cardiac cycle. It then estimates the onset of mechanical contraction (OMC) — essentially the phase angle at which that region begins to thicken. Collecting the OMC phase angles from all sampled regions yields the phase histogram: a distribution of contraction-onset timings across the whole ventricle.¹⁸

Electrical versus mechanical dyssynchrony

It is essential to distinguish two related but different ideas:

Electrical dyssynchrony is inferred from the surface ECG, principally QRS duration, and reflects the spread of electrical activation across the ventricles.
Mechanical dyssynchrony is the actual dispersion in the timing of regional myocardial contraction.

The two are correlated but not equivalent. Some patients with a wide QRS contract relatively synchronously, while some with a narrow QRS harbor meaningful mechanical dyssynchrony.¹³¹⁴ Phase analysis measures the mechanical timing directly, which is why it has been studied as a complement to, rather than a copy of, QRS-based selection.³⁴

For a grounding in how gated SPECT is acquired and quality-controlled before any phase processing is attempted, see our guides to cardiac SPECT MPI quality control and gated SPECT MPI and LVEF quantification.

Key Technical Principles

The Fourier phase model

For each sampled left-ventricular region, the regional count signal $C (t)$ varies periodically through the R-R interval $T$ . Phase analysis approximates this curve with a Fourier series. The first-harmonic case is the didactic one:¹⁸

C (t) \approx a_{0} + a_{1} cos (ω t) + b_{1} sin (ω t), ω = \frac{2 π}{T}

The phase angle of the region — its onset of mechanical contraction — is then recovered from the harmonic coefficients:

φ = atan2 (b_{1}, a_{1}), φ \in [0^{\circ}, 36 0^{\circ})

The full cardiac cycle maps to 360°, so a single frame of an $N$ -frame gate corresponds to $36 0^{\circ} / N$ : 45° for 8 frames and 22.5° for 16 frames. The continuous harmonic fit interpolates the onset timing between frames, so the effective temporal resolution of the phase estimate is far finer than the frame spacing when counts are adequate. Phantom work using realistic gated data has demonstrated multiharmonic temporal resolution on the order of a sixtieth of the cardiac cycle (roughly 5.6°).⁷ Commercial implementations such as the Emory Cardiac Toolbox use up to three harmonics rather than one.¹⁸

The four phase-histogram metrics

Once every region's phase angle is collected into the phase histogram $p (φ)$ , a handful of summary statistics describe its spread and shape. The phase standard deviation measures dispersion about the mean phase $\overset{φ}{ˉ}$ across the $M$ sampled regions:

PSD = \frac{1}{M} i = 1 \sum M (φ_{i} - \overset{φ}{ˉ})^{2}

Histogram bandwidth is the width of the band containing a fixed fraction of the distribution. The Emory Cardiac Toolbox defines bandwidth as the 95% span of the phase histogram:

BW = φ_{97.5%} - φ_{2.5%}

Phase entropy is the Shannon entropy of the normalized phase histogram, usually reported as a percentage of its maximum, where a higher value indicates a more disordered, less synchronous contraction:

E = - \frac{1}{lo g N} i = 1 \sum N p_{i} lo g p_{i} \times 100%

Here $p_{i}$ is the fraction of regions in histogram bin $i$ and $N$ is the number of occupied bins. The following table summarizes the metrics in common clinical use.¹⁸¹⁵

Phase metric	What it captures	Units	Direction with more dyssynchrony
Peak phase	Most common contraction-onset timing	degrees	shifts, less specific alone
Phase standard deviation (PSD)	Dispersion of onset timings about the mean	degrees	increases
Histogram bandwidth (BW)	95% span of the phase histogram	degrees	increases
Phase entropy	Disorder of the phase distribution	percent	increases

Bandwidth and phase standard deviation are the two metrics with the strongest validation against independent reference methods, and they are the ones most commonly reported.³

Why the numbers are software-specific

This is the single most important caveat in the whole field: phase-analysis normal limits and cut-points are specific to the software package, the normal database, and the acquisition protocol, and they are not interchangeable. A multi-software comparison built on a common normal database found that reported normal phase standard deviation differed roughly two-fold across programs, and that bandwidth and entropy also differed significantly; all of the metrics depend on sex, ejection fraction, and left ventricular volume.¹⁵

Software package	Normal phase SD (JSNM normal database)
Emory Cardiac Toolbox (ECTb)	11.5° ± 5.5°
Quantitative Gated SPECT (QGS)	5.3° ± 3.3°
Heart Function View (HFV)	5.4° ± 2.5°
cardioREPO	10.3° ± 3.2°

The practical consequence is unavoidable: a phase standard deviation of, say, 12° is near-normal for one package and clearly abnormal for another. A phase value is interpretable only against the normal database of the software that produced it.¹⁵¹⁸

For the reconstruction choices that feed these calculations, our overview of SPECT reconstruction with filtered back projection and OSEM explains how count statistics and filtering propagate into quantitative outputs.

Clinical Impact

Validation against independent reference methods

Phase analysis was not accepted on face value; it was cross-checked against other measures of dyssynchrony. Compared with tissue Doppler imaging, histogram bandwidth and phase standard deviation correlated strongly (bandwidth $r = 0.89$ , phase SD $r = 0.80$ ), while the higher-order shape metrics skewness and kurtosis correlated less well.³ A separate comparison against real-time three-dimensional echocardiography supported the same two metrics as the workhorses, with substantial dyssynchrony operationally defined in that Emory Cardiac Toolbox work as a bandwidth of at least 135° and a phase SD of at least 43°.⁵

CRT-response prediction

Several single-center studies examined whether baseline mechanical dyssynchrony predicts response to CRT. Using the Emory Cardiac Toolbox, responders had substantially larger baseline dyssynchrony than non-responders, with reported cut-points of a bandwidth of about 135° and a phase SD of about 43°.⁴ A separate study using Quantitative Gated SPECT reported its own, markedly different cut-points — a bandwidth near 72.5° and a phase SD near 19.6°.⁹ The two sets of numbers are not contradictory; they simply belong to different software and must never be mixed. This is the clinical face of the software-specificity rule.⁹¹⁵

Prognosis across populations

Beyond device selection, larger mechanical dyssynchrony has been associated with worse outcomes in several distinct populations:

In end-stage renal disease, a histogram bandwidth at or above the cohort median (about 62°) was associated with worse survival.¹¹
In ischemic LV dysfunction with a narrow QRS studied with gated PET, phase SD was an independent predictor of mortality.¹³
In non-ischemic cardiomyopathy with an ejection fraction of 35–50% and a QRS under 150 ms, the highest tertile of phase SD independently predicted all-cause mortality.¹⁴

Mechanical dyssynchrony is also common in the populations where it is sought: in one cohort with an ejection fraction at or below 35%, roughly half of patients met a significant-dyssynchrony threshold.¹² For how perfusion and function integrate into the broader cardiac workup, see Tc-99m sestamibi and tetrofosmin myocardial perfusion SPECT and quantitative myocardial blood flow with cardiac PET.

Practical Optimization Tips

Phase analysis is reproducible, but only if it is performed and reported with discipline.

Place the valve plane carefully. Reproducibility studies found that manual base-plane placement markedly improved repeatability compared with fully automated placement; in normal controls the mean absolute difference in phase SD fell from about 12° to about 1° when the base plane was set manually.⁶ Overall, interobserver and intraobserver reproducibility were excellent (intraclass correlation coefficients at or near 0.99–1.00), but repeatability is worse in dysfunctional ventricles than in normals.⁶
Protect count statistics. The Fourier fit degrades when regional counts are low. Ensure adequate injected activity, acquisition time, and energy-window settings so each sampled region carries enough counts for a stable harmonic fit.⁷
Prefer 16-frame gating when statistics allow, but do not panic over 8. The multiharmonic fit recovers sub-frame timing, so an 8-frame study is usable; frame count matters less than adequate counts, though 16 frames is generally preferred.⁷
Report sex-specific normals. Normal phase metrics differ between men and women; interpret against the sex-matched normal database.¹¹⁵
Never transfer cut-points between packages. State the software and normal database in the report, and compare every value to that package's own normals. A number without its software context is uninterpretable.¹⁵¹⁶
Treat the result as an adjunct. Phase metrics support, but do not replace, guideline-directed assessment and perfusion interpretation.¹⁶

Regulatory Considerations

Gated SPECT phase analysis is a post-processing adjunct to a myocardial perfusion study, so it inherits the regulatory and accreditation framework of nuclear cardiology rather than creating a new one. Several layers apply:

Imaging guidelines. The acquisition, processing, and interpretation of SPECT MPI follow the governing nuclear-cardiology imaging guidelines, which on the date of this article are the 2018 ASNC SPECT MPI guideline document covering instrumentation, acquisition, processing, and interpretation.¹⁶ The SNMMI/ASNC/SCCT cardiac SPECT/CT and PET/CT guideline provides the complementary procedure-level framework.¹⁷
Software and quantification. Phase-analysis outputs are produced by vendor-specific software with its own normal databases. Because the metrics are not interchangeable, the physicist and the laboratory should document which package and database are in use and ensure the normal limits match the clinical protocol.¹⁵
Radioactive material program. The underlying study uses byproduct material and is governed by 10 CFR Part 35 (or the equivalent Agreement State program) for authorized use, dose, and quality management, with 10 CFR Part 20 radiation protection limits in the background. Phase analysis adds no radioactive material and no additional dose.
Accreditation. Laboratories pursuing nuclear medicine or nuclear cardiology accreditation must document QC and processing procedures; the gated SPECT study underlying phase analysis is part of that documented program.

Of the states DRPS serves, Florida, Maryland, Virginia, California, Nevada, Pennsylvania, New York, and New Jersey are NRC Agreement States that administer their own medical-use programs, while Washington, DC and Delaware are regulated directly by the NRC for byproduct material. Always confirm the authority having jurisdiction. For the QC foundation that any quantitative SPECT output depends on, see cardiac SPECT MPI quality control.

Frequently Asked Questions (FAQs)

What is gated SPECT phase analysis?

Phase analysis is a post-processing technique applied to ECG-gated myocardial perfusion SPECT. It fits a Fourier function to each left-ventricular region's count-versus-time curve to estimate the onset of mechanical contraction, then plots those timings as a phase histogram whose spread quantifies mechanical dyssynchrony. It is derived from data already acquired for a routine gated study, so it adds no scan time, dose, or extra visit.¹

How is mechanical dyssynchrony different from electrical dyssynchrony?

Electrical dyssynchrony is inferred from the ECG, mainly QRS duration; mechanical dyssynchrony is the actual dispersion in regional contraction timing. The two are related but not identical, which is why direct mechanical measurement is of interest.³⁴

Are phase SD and bandwidth values comparable between software packages?

No. Reported normals can differ roughly two-fold between packages, so a value from one program cannot be compared with a value from another. Interpret every number against the normal database of the same software.¹⁵

Does 8-frame gating ruin phase analysis?

Not necessarily. The multiharmonic Fourier fit interpolates onset timing far below one frame, and phantom work shows fine temporal resolution when counts are adequate. Adequate counts matter more than frame number, though 16-frame gating is generally preferred.⁷

Can phase analysis predict who responds to cardiac resynchronization therapy?

It has shown value in several single-center studies, with larger baseline dyssynchrony associated with response, but the cut-points are software-specific and derive from small cohorts. It is supportive, not a standalone selection criterion.⁴⁹

What counts as a significant phase SD or bandwidth?

There is no universal threshold. In one Emory Cardiac Toolbox dataset, normal controls showed a phase SD near 9° and a bandwidth near 29°, while severe LV dysfunction showed values several times higher — but the thresholds depend on software, sex, ejection fraction, and LV volume.¹⁰¹⁵

Key Takeaways

Phase analysis extracts mechanical-timing information from a routine gated SPECT MPI study using a Fourier fit, with no added acquisition, dose, or patient visit.¹⁷
The phase histogram is summarized by phase standard deviation, bandwidth, and entropy; bandwidth and phase SD are the best-validated against independent methods.³
Mechanical dyssynchrony is not the same as electrical (QRS) dyssynchrony, which is the central reason to measure it directly.³⁴
Phase analysis is highly reproducible when the valve plane is set carefully and counts are adequate.⁶
Normal limits and cut-points are software-, database-, and protocol-specific and must never be transferred between packages.¹⁵
The technique is an adjunct interpreted under current imaging guidelines, using byproduct material governed by 10 CFR Part 35 or the Agreement State equivalent.¹⁶

Conclusion

Gated SPECT phase analysis is one of the rare quantitative tools in imaging that costs nothing extra to acquire: the timing information lives inside a study already performed for perfusion and function. Properly applied, it provides a reproducible, automated measurement of left ventricular mechanical dyssynchrony that correlates with independent reference methods and carries prognostic weight across several patient populations.

Its power comes with a discipline. The phase histogram metrics are only as meaningful as the normal database behind them, and those databases are not interchangeable across software. A phase standard deviation, a bandwidth, or an entropy value must always be reported with its software, its normal database, and the acquisition protocol that produced it. Read that way — as a context-bound, adjunctive measurement rather than a universal number — phase analysis is a genuinely useful addition to the gated SPECT report.

How DRPS Can Help

Diagnostic Radiation Physics Services supports nuclear cardiology and nuclear medicine programs across Florida, Maryland, Virginia, Washington DC, California, Nevada, Pennsylvania, New York, New Jersey, and Delaware with PET/CT and nuclear medicine physics support, gated SPECT acquisition and processing review, quantitative software and normal-database verification, QC program design, and accreditation support — all performed by board-certified medical physicists.

A strong quantitative nuclear cardiology program is not just about generating numbers; it is about generating numbers that mean the same thing every time, against the right normal database, so the report can be trusted. DRPS helps laboratories build that consistency into their medical physics consulting and QC workflow.

Related Resources

References

Chen J, Garcia EV, Folks RD, et al. Onset of left ventricular mechanical contraction as determined by phase analysis of ECG-gated myocardial perfusion SPECT imaging: development of a diagnostic tool for assessment of cardiac mechanical dyssynchrony. J Nucl Cardiol. 2005;12(6):687-695. doi:10.1016/j.nuclcard.2005.06.088. doi.org
Trimble MA, Borges-Neto S, Smallheiser S, et al. Evaluation of left ventricular mechanical dyssynchrony as determined by phase analysis of ECG-gated SPECT myocardial perfusion imaging in patients with left ventricular dysfunction and conduction disturbances. J Nucl Cardiol. 2007;14(3):298-307. doi:10.1016/j.nuclcard.2007.01.041. doi.org
Henneman MM, Chen J, Ypenburg C, et al. Phase analysis of gated myocardial perfusion single-photon emission computed tomography compared with tissue Doppler imaging for the assessment of left ventricular dyssynchrony. J Am Coll Cardiol. 2007;49(16):1708-1714. doi:10.1016/j.jacc.2007.01.063. doi.org
Henneman MM, Chen J, Dibbets-Schneider P, et al. Can LV dyssynchrony as assessed with phase analysis on gated myocardial perfusion SPECT predict response to CRT? J Nucl Med. 2007;48(7):1104-1111. doi:10.2967/jnumed.107.039925. doi.org
Marsan NA, Henneman MM, Chen J, et al. Real-time three-dimensional echocardiography as a novel approach to quantify left ventricular dyssynchrony: a comparison study with phase analysis of gated myocardial perfusion SPECT. J Am Soc Echocardiogr. 2008;21(7):801-807. doi:10.1016/j.echo.2007.12.006. doi.org
Trimble MA, Velazquez EJ, Adams GL, et al. Repeatability and reproducibility of phase analysis of gated SPECT myocardial perfusion imaging used to quantify cardiac dyssynchrony. Nucl Med Commun. 2008;29(4):374-381. doi:10.1097/MNM.0b013e3282f81380. doi.org
Chen J, Faber TL, Cooke CD, Garcia EV. Temporal resolution of multiharmonic phase analysis of ECG-gated myocardial perfusion SPECT studies. J Nucl Cardiol. 2008;15(3):383-391. doi:10.1016/j.nuclcard.2008.02.014. doi.org
Chen J, Garcia EV, Lerakis S, et al. Left ventricular mechanical dyssynchrony as assessed by phase analysis of ECG-gated SPECT myocardial perfusion imaging. Echocardiography. 2008;25(10):1186-1194. doi:10.1111/j.1540-8175.2008.00782.x. doi.org
Boogers MM, Van Kriekinge SD, Henneman MM, et al. Quantitative gated SPECT-derived phase analysis on gated myocardial perfusion SPECT detects left ventricular dyssynchrony and predicts response to cardiac resynchronization therapy. J Nucl Med. 2009;50(5):718-725. doi:10.2967/jnumed.108.060657. doi.org
Atchley AE, Trimble MA, Samad Z, et al. Use of phase analysis of gated SPECT perfusion imaging to quantify dyssynchrony in patients with mild-to-moderate left ventricular dysfunction. J Nucl Cardiol. 2009;16(6):888-894. doi:10.1007/s12350-009-9136-3. doi.org
AlJaroudi W, Aggarwal H, Venkataraman R, et al. Impact of left ventricular dyssynchrony by phase analysis on cardiovascular outcomes in patients with end-stage renal disease. J Nucl Cardiol. 2010;17(6):1058-1064. doi:10.1007/s12350-010-9271-x. doi.org
Samad Z, Atchley AE, Trimble MA, et al. Prevalence and predictors of mechanical dyssynchrony as defined by phase analysis in patients with left ventricular dysfunction undergoing gated SPECT myocardial perfusion imaging. J Nucl Cardiol. 2011;18(1):24-30. doi:10.1007/s12350-010-9310-7. doi.org
AlJaroudi W, Alraies MC, Hachamovitch R, et al. Association of left ventricular mechanical dyssynchrony with survival benefit from revascularization: a study of gated positron emission tomography in patients with ischemic LV dysfunction and narrow QRS. Eur J Nucl Med Mol Imaging. 2012;39(10):1581-1591. doi:10.1007/s00259-012-2171-3. doi.org
Goldberg AS, Alraies MC, Cerqueira MD, Jaber WA, AlJaroudi WA. Prognostic value of left ventricular mechanical dyssynchrony by phase analysis in patients with non-ischemic cardiomyopathy with ejection fraction 35-50% and QRS less than 150 ms. J Nucl Cardiol. 2014;21(1):57-66. doi:10.1007/s12350-013-9787-y. doi.org
Nakajima K, Okuda K, Matsuo S, et al. Comparison of phase dyssynchrony analysis using gated myocardial perfusion imaging with four software programs: based on the Japanese Society of Nuclear Medicine working group normal database. J Nucl Cardiol. 2017;24(2):611-621. doi:10.1007/s12350-015-0333-y. doi.org
Dorbala S, Ananthasubramaniam K, Armstrong IS, et al. Single photon emission computed tomography (SPECT) myocardial perfusion imaging guidelines: instrumentation, acquisition, processing, and interpretation. J Nucl Cardiol. 2018;25(5):1784-1846. doi:10.1007/s12350-018-1283-y. doi.org
Dorbala S, Di Carli MF, Delbeke D, et al. SNMMI/ASNC/SCCT guideline for cardiac SPECT/CT and PET/CT 1.0. J Nucl Med. 2013;54(8):1485-1507. doi:10.2967/jnumed.112.105155. doi.org
Nakajima K, Matsumoto N, Kasai T, et al. Normal values and standardization of parameters in nuclear cardiology: Japanese Society of Nuclear Medicine working group database. Ann Nucl Med. 2016;30(3):188-199. doi:10.1007/s12149-016-1065-z. doi.org