Cardiac MRI and nuclear perfusion imaging

01Purpose and principlesWhat the assessment is for and the core concepts behind it.

Cardiac MRI and nuclear perfusion answer overlapping but different questions. CMR can connect chamber size, global and regional function, myocardial oedema, fibrosis or infarction pattern, perfusion and vascular anatomy. Nuclear myocardial perfusion imaging, most commonly SPECT in routine UK practice, measures the relative distribution of a radiotracer under stress and rest conditions and can add gated ventricular function. Choosing between them depends on whether the decisive question is coronary physiology, myocardial tissue, ventricular structure, device or claustrophobia constraints, renal function, radiation, local expertise and the patient's ability to complete the protocol.

CMR builds the examination from sequences rather than a single image. Localisers establish anatomy. ECG-gated balanced steady-state free-precession cine stacks capture the cardiac cycle and support ventricular volumes, mass, ejection fraction and regional wall motion. T2-weighted or quantitative T2 approaches assess oedema; native T1, post-contrast T1 and extracellular-volume methods assess diffuse tissue change. Vasodilator stress first-pass imaging tracks gadolinium through myocardium. Late gadolinium enhancement, acquired after contrast has distributed, identifies regions with expanded extracellular space. The protocol must be tailored: a complete cardiomyopathy study differs from an ischaemia-only examination.

Perfusion imaging tests coronary flow reserve rather than merely showing a narrowing. A vasodilator creates a flow difference between myocardium supplied by normal and flow-limiting vessels. In stress CMR, a genuine inducible defect is typically darker than surrounding myocardium for several cardiac phases, begins in the subendocardium, respects a coronary territory and appears without matching scar. In SPECT, tracer uptake is relative: a region supplied by a restricted artery receives less tracer during stress than better-perfused myocardium. Rest imaging distinguishes reversible from persistent relative reduction.

Relative imaging has a central limitation. If all major coronary territories have similarly reduced hyperaemic flow, no region provides a normal reference and severe balanced disease can be underestimated. Clues include widespread symptoms or ECG change despite apparently uniform perfusion, transient ventricular dilation, a fall in post-stress function, diffuse subendocardial CMR hypoperfusion, or high-risk coronary anatomy already suspected. A normal-looking colour map cannot overrule strong discordant clinical evidence.

A technically correct report separates observation from interpretation. State whether stress was adequate, which segments are affected, whether defects are reversible or fixed, their extent and transmurality, ventricular volumes and function, scar or non-ischaemic enhancement pattern, artefacts and any incidental consequential finding. The clinical conclusion should explain whether the study supports inducible epicardial coronary ischaemia, prior infarction, microvascular dysfunction, cardiomyopathy or a non-diagnostic result.

Key points

CMR combines cine ventricular function, anatomy, oedema and parametric mapping, first-pass perfusion and late gadolinium enhancement in one examination.
Cine imaging quantifies volumes and ejection fraction without geometric assumptions, while regional wall motion may reveal ischaemia, infarction or cardiomyopathy.
After uncertain or non-diagnostic CT coronary angiography in stable chest pain, first-pass contrast-enhanced MR perfusion and myocardial perfusion scintigraphy with SPECT are alternative functional tests selected by patient factors and local expertise; they are not routinely sequential.
A reversible stress perfusion defect appears during stress and resolves at rest, supporting inducible ischaemia; match it to image quality, wall motion, coronary territory and scar.
SPECT perfusion compares relative tracer uptake. Reversible reduction supports ischaemia; a persistent reduction may be scar or artefact and should be checked against gated motion, attenuation-corrected images and raw data.
Balanced multivessel ischaemia can make relative perfusion look deceptively uniform; ventricular dilation, post-stress functional change and clinical risk can reveal the mismatch.
Dark-rim artefact on CMR and attenuation or extracardiac-activity artefact on SPECT can mimic defects; artefact recognition is part of interpretation, not an optional technical footnote.
CMR avoids ionising radiation and is strong for tissue characterisation; SPECT is widely available and provides established physiological assessment but uses radiation and has lower spatial resolution.
NICE lists stress MR perfusion and SPECT among functional tests after uncertain or non-diagnostic CT coronary angiography in stable chest pain.
Every report should answer the referral question, state stress adequacy and image limitations, and recommend the next step only where the imaging result supports it.

02Indications, selection and cautionsWhen it is useful, when urgency changes and important limitations.

Inducible ischaemia on stress CMR

A convincing defect develops during stress, persists beyond the initial arrival of contrast, is usually subendocardial, follows one or more coronary territories and exceeds any matching dark-rim artefact. Absence of late enhancement in that segment indicates viable myocardium at risk rather than established replacement scar. Diffuse circumferential subendocardial hypoperfusion may suggest multivessel epicardial disease or microvascular dysfunction and requires clinical correlation.

Ischaemic late-enhancement pattern

Infarction begins in the subendocardium because it is most vulnerable to reduced perfusion, then can extend transmurally. Enhancement that includes the subendocardium and follows a vascular territory supports ischaemic injury. Greater transmural extent implies less likelihood of functional recovery, but treatment decisions integrate wall thickness, function, symptoms, coronary anatomy and the amount of jeopardised viable myocardium.

Non-ischaemic myocardial injury

Mid-wall linear enhancement, subepicardial enhancement or patchy enhancement that does not respect a coronary territory directs attention toward myocarditis or cardiomyopathy rather than infarction. Oedema and mapping abnormalities help identify active injury, but values depend on scanner, sequence and local normal ranges. A pattern narrows the differential; it rarely supplies a complete aetiological diagnosis alone.

Reversible SPECT perfusion defect

Reduced tracer uptake at stress with normalisation or improvement at rest supports inducible ischaemia. Confirm the finding in orthogonal planes, evaluate its vascular distribution and extent, and inspect gated wall thickening, raw rotating projections and attenuation-corrected data. A coronary-territory defect with matching functional change is more persuasive than an isolated mild count difference.

Fixed SPECT defect

A reduction present at stress and rest may represent infarct scar, attenuation or less commonly severely reduced viable perfusion. Preserved wall motion and thickening in a typical attenuation location favour artefact, while matching regional dysfunction supports myocardial damage. Neither interpretation should be made without reviewing raw data and technical correction.

Balanced ischaemia and high-risk ancillary signsRed flag

Because SPECT displays relative uptake, homogeneous global flow reduction can look uniform. Transient ischaemic dilation, stress-induced left-ventricular dysfunction, extensive coronary calcification on hybrid imaging, symptoms or ECG change and a high-risk clinical profile should lower confidence in an apparently normal perfusion distribution.

CMR dark-rim and SPECT attenuation artefacts

CMR dark-rim artefact often appears for only a few frames, may be one pixel thick and can extend beyond a single vascular territory. Breast or diaphragmatic attenuation on SPECT produces characteristic regional reductions, sometimes with preserved gated motion. Motion, mistriggering and extracardiac tracer activity create additional patterns; raw-data review is essential.

Red flags requiring action

Current chest pain with instability, dynamic ECG change or rising troponin: manage as possible acute coronary syndrome before elective stress imaging.
Decompensated heart failure, uncontrolled important arrhythmia, severe hypotension or another contraindication to the proposed stress agent: postpone stress and stabilise or choose another route.
A non-MR-conditional implant, retained ferromagnetic foreign body or uncertain device status: stop and complete formal MRI-safety assessment rather than relying on patient recollection alone.
A new severe perfusion abnormality, stress-induced wall-motion abnormality or high-risk ventricular arrhythmia during testing: terminate stress, treat clinically and communicate urgently.
A fixed defect with acute symptoms or a new regional wall-motion abnormality is not automatically old scar; reconcile timing, ECG, biomarkers and artefact before closing the diagnosis.
Pregnancy or possible pregnancy requires an explicit justification and radiation-safety pathway before radionuclide imaging.

03Method and interpretationA systematic approach to the test and its findings.

Investigation order

Read from the initial assessment onwards. Tests may run in parallel in urgent care; first-line, preferred, confirmatory, definitive and gold-standard labels appear only when the chapter explicitly states them.

01
Vasodilator stress perfusion CMRFirst stepPreferred
Why
Serve as a NICE-listed functional imaging option after uncertain or non-diagnostic CT coronary angiography; it may be preferred when myocardial tissue and scar information plus avoidance of ionising radiation are valuable.
Interpretation and limitations
Confirm appropriate preparation, safety screening and an adequate physiological response. Compare stress and rest first-pass images, then match any defect to coronary distribution, late enhancement and wall motion. Record artefact or incomplete coverage and call a study non-diagnostic when they prevent a confident answer.
02
Comprehensive cardiac MRI
Why
Quantify biventricular size and function and characterise myocardium in suspected cardiomyopathy, myocarditis, infiltrative disease, congenital disease or an unexplained structural abnormality.
Interpretation and limitations
Use contiguous cine stacks for volumes and function, targeted anatomical planes, tissue-sensitive sequences, mapping with site-specific reference ranges and late enhancement where contrast is appropriate. Integrate pattern, distribution and clinical timing; do not label aetiology from an isolated mapping value.
03
Late gadolinium enhancement imaging
Why
Identify focal myocardial fibrosis, infarction or replacement injury and describe its distribution and transmural extent.
Interpretation and limitations
Null normal myocardium correctly, examine orthogonal planes and distinguish true enhancement from poor inversion time, motion and blood-pool artefact. Subendocardial coronary-territory enhancement supports ischaemic injury; mid-wall, subepicardial and patchy patterns support a non-ischaemic process.
04
Rest and stress SPECT myocardial perfusion imagingPreferred
Why
Serve as a NICE-listed functional imaging option after uncertain or non-diagnostic CT coronary angiography; it may be preferred when local expertise, availability or MRI-related patient factors favour radionuclide imaging, with gated function where acquisition permits.
Interpretation and limitations
Compare matched stress and rest slices and polar maps, quantify extent and severity in context, and review gated motion, attenuation-corrected and uncorrected images, raw projections and stress response. Separate reversible, fixed and equivocal defects and consider balanced disease when ancillary findings conflict.
05
ECG-gated SPECT function
Why
Add left-ventricular volume, ejection fraction, regional motion and systolic thickening to perfusion interpretation.
Interpretation and limitations
Use gated findings to test whether a fixed defect behaves like scar or attenuation. Arrhythmia, poor counts and gating errors can corrupt volumes or motion; always inspect beat histograms or quality indicators rather than accepting the calculated ejection fraction uncritically.
06
PET myocardial perfusion where available
Why
Measure perfusion with higher count resolution and, for selected tracers and protocols, quantify absolute myocardial blood flow and flow reserve.
Interpretation and limitations
Absolute flow can expose global or microvascular impairment that relative SPECT misses, but results depend on tracer, acquisition, kinetic model and local thresholds. Availability and radiation still influence selection.
07
Coronary CT or invasive angiography after functional imaging
Why
Define coronary anatomy when function shows consequential ischaemia, remains inconclusive or conflicts with a high-risk clinical picture.
Interpretation and limitations
Select the anatomical next test according to presentation and prior imaging. Functional abnormality does not identify every lesion morphology; anatomy and physiology are complementary, and invasive assessment is reserved for cases where it will guide treatment.

04Clinical next stepsHow the result changes management or prompts escalation.

01Stable chest-pain pathwayUse functional imaging after uncertain anatomical testingFirst stepA stable patient has symptoms compatible with angina and CT coronary angiography is uncertain or non-diagnostic for the functional significance of disease.

1Recheck symptom pattern, CT image quality, lesion location and the probability that the uncertain segment could explain the presentation.
2Choose one non-invasive functional test according to local expertise and patient factors. NICE options are myocardial perfusion scintigraphy with SPECT, stress echocardiography, first-pass contrast-enhanced MR perfusion, or MR imaging for stress-induced wall-motion abnormalities.
3Ensure the patient can safely receive the selected stress agent and complete the examination, including MRI and contrast screening where relevant.
4Interpret perfusion with stress adequacy, ventricular function, scar and artefact, and identify whether the abnormality is limited, extensive, fixed or non-diagnostic.
5Consider invasive coronary angiography when functional imaging remains inconclusive and the result will determine revascularisation or another consequential management decision.

02Cardiomyopathy pathwayBuild tissue characterisation around the suspected mechanismEchocardiography or clinical findings show unexplained ventricular dysfunction, hypertrophy, regional abnormality or inflammatory features.

1Define the clinical question and screen for instability, device safety, renal status and ability to breath-hold before selecting the protocol.
2Acquire biventricular cine volumes and function, then add oedema-sensitive imaging, native mapping, perfusion or flow sequences according to the phenotype.
3Use late enhancement when appropriate to distinguish coronary-territory subendocardial injury from mid-wall, subepicardial or patchy non-ischaemic injury.
4Integrate CMR with ECG, biomarkers, family history, coronary assessment and extracardiac disease rather than treating an imaging pattern as a genetic or inflammatory diagnosis.
5Use the result to guide specialist referral, disease-specific work-up and a follow-up plan based on symptoms, ventricular trajectory and arrhythmic risk.

03Nuclear perfusion quality pathwayResolve a possible defect before assigning diseaseSPECT shows a mild fixed or reversible defect, or the perfusion map conflicts with symptoms, ECG or ventricular function.

1Inspect rotating raw projections for movement, truncation, body habitus and extracardiac activity, and verify correct axis reconstruction.
2Compare attenuation-corrected and uncorrected data because correction can remove a true attenuation pattern but can also introduce misregistration artefact.
3Review gated motion and thickening: preserved contraction within a fixed count defect favours attenuation, whereas matching dysfunction supports scar.
4Check stress adequacy, symptoms and ECG response, and look for transient dilation or post-stress functional change that may signal extensive disease.
5AlternativeReport an equivocal or non-diagnostic study honestly and recommend an alternative anatomical or functional test only if it will resolve a management decision.

04Scar and viability pathwaySeparate infarction, viable risk and non-ischaemic fibrosisLeft-ventricular dysfunction or a regional abnormality raises the question of prior infarction, inducible ischaemia or recoverable myocardium.

1Map regional wall thickness and motion on cine imaging or gated SPECT before interpreting perfusion or enhancement.
2Identify inducible perfusion abnormality and determine whether the same segment contains fixed scar or late enhancement.
3Describe the coronary distribution and transmural extent of scar; preserved unscarred or partially scarred myocardium may remain viable but does not guarantee clinical recovery.
4Relate viability and ischaemia burden to symptoms, coronary anatomy, procedural feasibility and overall risk in a multidisciplinary decision rather than using a binary scan label.

05Risks, monitoring and follow-upComplications, safety checks and further assessment.

During pharmacological stress, monitor symptoms, rhythm, heart rate and blood pressure according to the protocol, with trained staff and emergency equipment available.
Record whether caffeine or interacting medicines affected vasodilator preparation and whether the physiological response was adequate; inadequate stress can produce a falsely reassuring examination.
For serial CMR, compare ventricular volumes, ejection fraction, mass and tissue findings with the same segmentation and acquisition conventions where possible.
Do not schedule routine repeat perfusion solely because a prior study was abnormal. Repeat when symptoms, risk or treatment have changed and the result will alter management.
After a non-diagnostic study, document the exact cause—motion, arrhythmia, attenuation, poor counts, stress failure, contrast timing or incomplete coverage—so the next modality solves rather than repeats it.
Communicate unexpected severe dysfunction, extensive ischaemia, active-appearing myocardial injury or a dangerous extracardiac finding promptly rather than leaving it only in the final report.
Track cumulative radiation exposure conceptually when choosing repeated radionuclide or CT examinations, while prioritising the test needed to answer a consequential clinical question.

06Special situationsVariants, exceptions and circumstances that change the usual approach.

Function, perfusion and tissue form a causal sequence

Coronary flow limitation first impairs hyperaemic perfusion, then can produce stress wall-motion abnormality; prolonged injury creates necrosis and later scar. A single examination becomes more persuasive when the perfusion defect, regional function and enhancement pattern occupy the same coronary territory.

Late enhancement is distributional evidence

Gadolinium accumulates where extracellular space is expanded, so enhancement is not synonymous with infarction. The subendocardial coronary-territory pattern makes injury ischaemic; mid-wall or subepicardial distribution changes the mechanism under consideration.

Relative normality can conceal global abnormality

SPECT compares each region with the brightest myocardium. If every territory is underperfused, the differences shrink. High-risk clinical context, stress response and ancillary ventricular signs must therefore remain visible to the interpreter.

Gated motion adjudicates many fixed defects

Soft-tissue attenuation reduces counts without damaging myocardium. Preserved systolic thickening in the same segment makes attenuation more likely; true transmural scar more often has matching thinning or regional dysfunction, although small or non-transmural scars can retain motion.

A dark rim is judged across space and time

An artefactual CMR subendocardial rim often appears at contrast arrival, is thin, transient and may not follow one artery. A real defect persists through several frames, is wider, has a vascular distribution and gains credibility from matching symptoms or coronary anatomy.

Mapping requires local reference ranges

Native T1, T2 and extracellular volume vary with field strength, sequence and analysis method. The pattern and clinical setting matter more than comparing a patient's number with a threshold copied from another scanner.

A normal perfusion study answers a defined question

It lowers the likelihood of flow-limiting epicardial disease only to the extent that stress was adequate and images were diagnostic. It does not exclude plaque, vasospasm, intermittent thrombosis or every form of microvascular dysfunction.

07Common pitfallsFrequent interpretation and management errors.

01
Sending a patient with active suspected acute coronary syndrome for routine stress imaging before urgent ECG, biomarker and clinical assessment.
02
Calling any late gadolinium enhancement an infarct without checking whether it involves the subendocardium and follows a coronary territory.
03
Diagnosing myocarditis from an isolated raised mapping value without regional pattern, oedema, clinical timing or local reference ranges.
04
Treating a brief one-pixel CMR dark rim as inducible ischaemia without examining persistence and vascular distribution.
05
Calling a fixed inferior or anterior SPECT defect scar before reviewing gated wall thickening, raw data and attenuation correction.
06
Accepting a visually uniform SPECT study as low risk despite inadequate stress, transient ventricular dilation, post-stress dysfunction or a strongly discordant clinical picture.
07
Ignoring arrhythmia-related gating error when interpreting calculated ejection fraction or apparent regional motion.
08
Ordering gadolinium-enhanced MRI or radionuclide imaging without the required renal, pregnancy, device and agent-specific safety checks.
09
Repeating the same technically failed modality without stating why it failed and how the next examination will answer the question.
10
Recommending revascularisation from an imaging label alone without integrating symptoms, anatomy, ischaemic extent, scar and procedural suitability.

Find your next topic