CT myocardial perfusion

The tissue-level question

CTCA shows the pipe; CT perfusion shows the myocardium it supplies

Myocardial CT perfusion (CTP) evaluates contrast delivery through the coronary and microvascular circulation into myocardium, usually at rest and during vasodilator stress. The core diagnostic target is inducible ischaemia.

Perfusion can resolve equivocal intermediate stenosis, dense-calcium overestimation, serial/multivessel disease, and situations where microvascular dysfunction or prior infarction contributes to symptoms. It adds acquisition, contrast, radiation, pharmacological stress, and operational complexity.

Why stress unmasks disease

Rest flow can be preserved until coronary reserve is challenged

Autoregulation maintains resting myocardial flow across a range of epicardial stenoses. Vasodilator stress lowers microvascular resistance and raises expected flow. A flow-limiting epicardial lesion—or abnormal microvascular reserve—prevents the normal increase, producing a relative or absolute stress perfusion deficit.

• Rest normal, stress defect: reversible ischaemia.
• Rest defect, similar stress defect: fixed defect, often infarction/scar, though artifact must be excluded.
• Rest defect, larger at stress: peri-infarct ischaemia.
• Globally reduced stress flow without focal epicardial correlate: consider diffuse CAD, microvascular dysfunction, inadequate stress, or technical error.

Choose it for an unresolved question

High-value clinical settings

• Moderate (50–69%) or selected severe-appearing stenosis whose haemodynamic relevance is uncertain.
• Dense calcified or mixed plaque where blooming may overestimate luminal narrowing.
• Stents or prior revascularisation when anatomy is limited but myocardial territory can still be assessed.
• Serial lesions or multivessel disease requiring territory-level ischaemia assessment.
• Suspected ischaemia with nonobstructive coronary arteries (INOCA), recognising that quantitative/dynamic methods are more informative for global microvascular dysfunction.
• When CT-FFR is unavailable, rejected, anatomically discordant, or unsuitable—and local CTP expertise exists.

CTP is not a routine add-on to every CTCA. It should be planned when the result can change management and when patient, scanner, staff, contrast, and radiation considerations are acceptable.

One stress snapshot

Static CTP: relative myocardial attenuation at peak stress

Static CTP acquires one ECG-gated dataset during first-pass myocardial enhancement at pharmacological hyperaemia. Hypoperfused myocardium enhances less than normally perfused myocardium.

Dual-energy/spectral iodine maps can improve contrast and reduce some beam-hardening problems, but they remain dependent on correct acquisition and artifact recognition.

Time–attenuation sampling

Dynamic CTP: serial imaging for myocardial blood-flow estimation

Dynamic CTP repeatedly samples the myocardium during contrast passage, producing arterial input and tissue time–attenuation curves. Validated models derive myocardial blood flow (MBF) and related parameters.

• Advantages: quantitative/semiquantitative flow, detection of balanced multivessel disease, assessment of global reduction, and potential microvascular insight.
• Requirements: scanner with sufficient cardiac coverage/temporal strategy, stable breath-hold, robust motion correction, and vendor-specific validated processing.
• Costs: generally higher radiation than static CTP, multiple phases, more complex reconstruction, and model-dependent thresholds.
• Threshold caution: absolute MBF cut-points differ across scanner, stress agent, acquisition, and software; use locally validated ranges rather than importing one universal value.

From contrast passage to a flow map

Dynamic CTP acquisition and analysis workflow

1. Create hyperaemia. Confirm that vasodilator exposure and haemodynamic/symptom response are adequate before interpreting low flow.
2. Sample the first pass. Repeated ECG-triggered datasets track the contrast bolus through the arterial blood pool and myocardium. One multicentre protocol acquired approximately 10-15 samples over about 30 seconds; this is an example, not a universal prescription.
3. Register the myocardium. Respiratory and cardiac motion correction must align the LV wall across time. Registration failure can create false regional hypoperfusion.
4. Define the arterial input. A blood-pool region, commonly the descending aorta or LV cavity depending on software, supplies the arterial input function.
5. Fit a validated model. Deconvolution, compartmental, upslope, or hybrid approaches estimate MBF. Values from different models and vendors are not interchangeable.
6. Display segmental and pixel maps. Review the short-axis series, 17-segment polar map, curves, and numerical values together. Avoid relying on colour alone.

Multicentre evidence

What dynamic CTP adds after coronary CTA

The 2022 international multicentre SPECIFIC study enrolled 132 patients at nine centres; 114 completed coronary CTA, adenosine-stress dynamic CT perfusion, invasive coronary angiography, and the study pathway. The vessel-level reference combined invasive FFR and angiographic severity. Of 289 evaluated vessels, 74 (26%) had haemodynamically significant disease.

Adding perfusion reduced false-positive anatomical calls and improved overall classification, at the cost of some sensitivity. Absolute MBF achieved an area under the ROC curve of 0.79; relative MBF achieved 0.82. Median dose-length product was 313 mGy·cm for CT perfusion and 138 mGy·cm for CTA in this protocol.

Design the sequence deliberately

Rest-first and stress-first strategies

The specific bolus, scan delay, ECG phase, kVp, coverage, and rest-stress interval are scanner- and agent-specific. Use a validated departmental protocol. Minimise total contrast and dose while preserving the diagnostic question.

Protocol details that determine validity

• Caffeine and methylxanthines: can blunt vasodilator stress; follow agent-specific withholding instructions and document uncertainty.
• Nitrates and beta-blockers: may be needed for CTA but can influence haemodynamics and sequence design. Record what was given and when.
• Acquisition phase: systolic imaging may reduce motion in higher heart rates and offers greater myocardial thickness; diastolic imaging may align better with coronary CTA. Use the phase validated for the platform.
• Radiation control: narrow z-axis coverage when appropriate, prospective ECG triggering, low kVp in suitable patients, iterative reconstruction, and omission of non-contributory phases.
• Post-processing: automated output still requires checking contours, arterial input, motion registration, curves, and segment assignment.

Stress is a clinical procedure

Screen, monitor, and prepare to reverse

Vasodilators used in CTP include adenosine, regadenoson, and dipyridamole. A commonly used adenosine protocol is 140 µg/kg/min, but dosing/timing must follow the validated local protocol and supervising clinician.

1. Confirm indication, consent, IV access, renal/contrast considerations, and fasting/caffeine/methylxanthine instructions.
2. Screen for active ACS/instability, severe hypotension, high-grade AV block without appropriate pacing, significant bronchospastic disease, severe uncontrolled hypertension, recent dipyridamole, hypersensitivity, and other agent-specific contraindications.
3. Record baseline symptoms, blood pressure, heart rate/rhythm, and ECG.
4. Maintain continuous clinical/ECG monitoring and repeat haemodynamics during and after stress.
5. Keep oxygen, resuscitation equipment, trained staff, and the protocol-specific reversal agent (commonly aminophylline for persistent vasodilator effects) immediately available.
6. Stop/reverse according to symptoms, ECG, haemodynamics, or local emergency criteria.

Read myocardium systematically

A perfusion interpretation sequence

1. Confirm stress adequacy, contrast timing, motion correction, and global image quality.
2. Review standard short-axis plus 2-, 3-, and 4-chamber reformats at consistent thickness/window.
3. Look for a subendocardial defect extending toward epicardium in a coronary distribution and across contiguous slices.
4. Compare stress with rest at matched levels and phases.
5. Assess transmural extent, segment count, territory, and whether the defect is focal, multiterritorial, or global.
6. Review wall thickness, motion/function if available, fat, calcification, delayed enhancement, and other infarct clues.
7. Correlate with the actual patient-specific coronary anatomy, dominance, stenoses, stents/grafts, and CT-FFR.
8. Interrogate raw/source images and iodine/MBF maps for artifacts before calling ischaemia.

Stress–rest logic

Reversible, fixed, and peri-infarct defects

A fixed defect is not “normal”; it is simply not reversible ischaemia. Report infarction, wall thinning, aneurysm, fat/calcification, or dysfunction in the impression when present.

Correlate, do not force

Perfusion territories vary with dominance and branch anatomy

Use the AHA 17-segment model for communication, then map the defect to the patient’s real coronary tree. LAD usually supplies anterior/anteroseptal and much of the apex; LCx the lateral wall; RCA the inferior/inferoseptal wall in right dominance. Boundaries overlap, and the apical cap is particularly variable.

• A defect crossing “textbook” territories may reflect left dominance, wraparound LAD, large ramus/OM, multivessel disease, or artifact.
• A small diagonal/OM lesion should match its actual branch territory, not the entire parent-vessel map.
• Grafts and collateral pathways change the relationship between native stenosis and perfused myocardium.

False defects are common

Recognise perfusion mimics before diagnosing ischaemia

True defects usually follow subendocardium, persist across contiguous slices, form a plausible coronary territory, and correlate with anatomy. None of these alone is sufficient.

Numbers need quality control

Dynamic MBF maps can expose global and balanced disease

Absolute or semiquantitative MBF can reveal globally reduced hyperaemic flow when relative static images appear deceptively uniform. It can also quantify regional severity and help investigate microvascular dysfunction.

• Inspect time–attenuation curves and motion-correction registration—not only the colour map.
• Confirm adequate stress response and arterial input sampling.
• Use scanner/software/agent-specific normal and abnormal ranges.
• Recognise that high resting flow can reduce apparent reserve even when stress flow is preserved.
• Global reduction requires a differential including multivessel epicardial CAD, diffuse disease, microvascular dysfunction, inadequate stress, and technical failure.

A practical pattern matrix

Cross-modality calibration

Quantitative perfusion CMR clarifies the meaning of flow and reserve

Quantitative first-pass CMR and dynamic CTP ask the same physiological question with different acquisition and modelling techniques: how much blood reaches myocardium at rest and during hyperaemia? CMR adds tissue characterisation and avoids ionising radiation; CT combines perfusion with high-resolution coronary anatomy and is often faster and more available.

The reviewed CMR literature reports approximate normal ranges of 0.8-1.3 mL/min/g for resting MBF, 2.3-3.7 mL/min/g for stress MBF, and 2.6-4.1 for MPR. These are orientation ranges, not thresholds for CT and not universal CMR cut-points. One cited CMR study used stress MBF at least 2.25 mL/min/g to distinguish normal from abnormal perfusion and values at or below 1.82 mL/min/g to help separate obstructive three-vessel disease from microvascular dysfunction; those research cut-points must remain technique-specific.

Beyond obstructive CAD

Quantitative perfusion can support evaluation of coronary microvascular dysfunction and may provide prognostic information. CMR has additional published applications in hypertrophic cardiomyopathy, amyloidosis, non-ischaemic cardiomyopathy, valvular disease, and cardiac allograft vasculopathy. These uses broaden the physiological framework but do not make CT and CMR interchangeable.

Complementary, not competing

CT perfusion versus CT-FFR

Select the test that answers the unresolved question with the least added risk and complexity. In some complex patients, anatomy, CT-FFR, and CTP are complementary; discordance should trigger careful source review rather than automatic trust in one output.

Structured physiology

Apply the CAD-RADS I modifier correctly

• I+: reversible defect or peri-infarct ischaemia.
• I−: no reversible defect; also used for a fixed prior-infarct defect without inducible ischaemia, with infarct described separately.
• I±: borderline/equivocal finding or unresolved anatomy–perfusion discordance.
• N: perfusion analysis itself is non-diagnostic.

If a defect lacks a concordant anatomical lesion, recheck both studies. If confidently artifactual, classify I−; if uncertainty remains, I±. Never manufacture an epicardial culprit to fit a perfusion map.

Clinical handoff

Minimum CTP report

1. State indication, stress agent, static/dynamic technique, rest–stress order, scanner/processing method, contrast/dose, and adequacy.
2. Document symptoms, ECG/hemodynamic response, adverse events, and reversal if used.
3. Describe image quality and artifacts.
4. Report defect location by AHA segments, coronary territory, extent, transmurality, and reversibility.
5. For dynamic CTP, report validated MBF values/units, reference ranges, and global/regional pattern.
6. Correlate with CTCA lesion, dominance, stent/graft, and CT-FFR if available.
7. State ischaemia present/absent/equivocal, fixed infarct separately, and CAD-RADS I result.

Integrate anatomy and tissue

Worked perfusion cases

Case A · 60% proximal LAD, stress-only anterior defect

Reversible LAD-territory ischaemia: I+. Confirm contiguous subendocardial pattern and exclude motion/beam hardening.

Case B · 75% calcified RCA, normal stress and rest perfusion

No inducible perfusion abnormality: I−. Reassess whether blooming overestimated stenosis and integrate symptoms; a negative CTP does not erase plaque burden.

Case C · Inferolateral defect identical at rest and stress with wall thinning

Fixed infarct pattern without inducible ischaemia: I−, but explicitly report prior infarct/scar and ventricular consequences.

Case D · Small fixed inferolateral core with larger stress defect

Peri-infarct ischaemia: I+. Report fixed and reversible extents separately.

Case E · No obstructive epicardial lesion; globally low dynamic stress MBF

Consider microvascular dysfunction, diffuse disease, inadequate vasodilator response, and technical/model error. Do not assign a focal culprit without anatomical support.

Case F · Three-vessel disease; static stress looks uniformly normal, dynamic MBF globally reduced

Balanced ischaemia can defeat relative static comparison. Validate stress and curves, then report multivessel/global reduction with anatomical correlation.

Active recall

Self-test: answer before opening

Printable quick revision

One-page CT perfusion card

• Static: one relative stress snapshot; lower dose, timing-sensitive.
• Dynamic: serial time curves → MBF; detects global/balanced reduction; more dose/complexity.
• Reversible: stress+ / rest− = ischaemia = I+.
• Fixed: stress+ / rest+ = infarct/no reversible ischaemia = I−; report scar.
• Peri-infarct: stress defect larger than rest = I+.
• Artifacts: beam hardening · motion · timing · partial volume · metal · misregistration.
• CTP vs CT-FFR: myocardium/flow vs epicardial pressure; extra stress scan vs no extra acquisition.
• Safety: screen · ECG/BP monitor · emergency readiness · reversal protocol.
• Final rule: a colour map is not a diagnosis until source images, stress adequacy, rest comparison, and coronary anatomy agree.

Provenance

Sources and scope

This chapter follows Dr Bhavana Reddy’s Webinar 4 CT myocardial perfusion teaching and discussion, approximately 01:28:16–01:41:55 plus Q&A. Automated captions contain transcription errors; safety, quantitative methods, evidence, and reporting principles were checked against the sources below.

1. Sliwicka O, et al. Dynamic myocardial CT perfusion imaging - state of the art. European Radiology. 2023;33:5509-5525. - physiology, clinical applications, acquisition, reconstruction, quantitative analysis, evidence, limitations, and future directions.
2. Nous FMA, et al. Dynamic Myocardial Perfusion CT for the Detection of Hemodynamically Significant Coronary Artery Disease. JACC: Cardiovascular Imaging. 2022;15:75-87. - international multicentre diagnostic-performance evidence and quantitative MBF methods.
3. Catania R, et al. Quantitative Stress First-Pass Perfusion Cardiac MRI: State of the Art. RadioGraphics. 2025;45:e240115. - MBF/MPR physiology, quantitative CMR technique, reference ranges, interpretation, artifacts, and clinical applications.
4. SCCT expert consensus on myocardial CT perfusion imaging - indications, static/dynamic acquisition, stress safety, interpretation, and reporting.
5. CAD-RADS 2.0 consensus - stress CTP integration and I+/I-/I± rules.
6. SCCT CTP consensus resource - clinical implementation and safety overview.
7. 2021 AHA/ACC multisociety chest-pain guideline executive summary - stress-testing contraindications and patient-selection context.

Educational material only. Stress CTP requires trained supervision, local medication/emergency protocols, scanner-specific acquisition, validated processing, and patient-specific clinical judgement. External references require internet; the chapter itself works offline.