MRA over-reads stenosis by design — turbulence causes signal loss that looks like narrowing, and the technique determines how much.
Orient first
- MRA does not image the lumen directly; it images FLOW or contrast within it. Anything that disturbs flow — turbulence at a stenosis, a tight curve, in-plane flow — causes signal loss that MIMICS or EXAGGERATES narrowing. Systematic overestimation of stenosis is a property of the technique, not an error.
- The techniques behave differently. TIME-OF-FLIGHT is flow-dependent and most prone to overestimation and to in-plane saturation; CONTRAST-ENHANCED MRA is faster, less flow-dependent and generally more accurate; phase-contrast additionally gives direction and velocity.
- A SUSCEPTIBILITY artefact from a clip, stent or coil creates a signal void that is indistinguishable from occlusion. Knowing what hardware is present is part of reading the study.
- TIMING is everything in contrast-enhanced MRA: too early and the artery is underfilled, too late and venous contamination obscures it.
- Always review the SOURCE images. Maximum-intensity projections hide low-signal plaque, overlying structures and small dissections, and can create an appearance of occlusion from a projection artefact.
Acquire the study
- Establish which technique was used, and whether contrast was given — the read differs substantially between them.
- Confirm the arterial timing by looking for venous contamination before assessing anything.
- Ensure the field of view covers the inflow and outflow of the segment in question, since a tandem lesion elsewhere changes management.
- Confirm which sequence you are on before judging any signal — see the MRI sequence primer.
- Establish what metallic hardware is present before interpreting any signal void.
The manoeuvre
- Assess technical adequacy: timing, venous contamination, motion and coverage.
- Review the SOURCE images first and use projections only for overview.
- Follow each vessel along its length and describe any narrowing by its location, length and severity.
- At any apparent stenosis, look for the flow-related signal void that indicates turbulence, and state that severity may be overestimated.
- Distinguish occlusion from a signal void caused by hardware, slow flow or in-plane saturation, using the other sequences.
- Assess for dissection: an intimal flap, a false lumen, or an intramural haematoma bright on fat-suppressed T1.
- Assess for aneurysm, measuring outer wall to outer wall perpendicular to the vessel axis.
- Assess collateral vessels, which indicate the chronicity and haemodynamic significance of a lesion.
- Correlate with any non-angiographic sequences in the study for the wall and surrounding tissue.
What confirms it
- A stenosis is a narrowing reproducible on the source images in more than one plane, with a consistent appearance across techniques.
What licenses you to exclude it
- ⚠️ MRA cannot reliably exclude a small aneurysm or a subtle dissection, and it does not grade a stenosis as accurately as catheter angiography. Say what would settle it — CT angiography or catheter angiography — rather than issuing a confident grade.
- A signal void adjacent to hardware excludes nothing; report the segment as non-assessable.
- A poorly-timed acquisition cannot exclude disease; report it as non-diagnostic for the arterial question.
The classic misread
- Grading a stenosis from projections alone.
- Calling occlusion where the signal void is caused by a clip or a stent.
- Reading a venously contaminated acquisition as an arterial study.
- Not stating that the technique overestimates severity.