Where the scar sits in the wall is the diagnosis: subendocardial or transmural in a coronary territory is infarction; mid-wall or epicardial, sparing the subendocardium, is non-ischaemic — then match the pattern to the cardiomyopathy.
Orient first
- Infarcts spread from the SUBENDOCARDIUM outward, in a coronary territory. Non-ischaemic scar does not start at the subendocardium.
- Late gadolinium enhancement needs a correct inversion time: normal myocardium must be nulled (black); a poorly nulled image hides or fakes scar.
- Report by the 17-segment AHA model, with the transmural extent of scar — it predicts whether a segment will recover after revascularisation.
Acquire the study
- Cine SSFP (short-axis stack and long axes) for volumes and function; T2-weighted or T2 mapping for oedema; native T1 and ECV where available; late gadolinium enhancement 10–15 min after contrast, in short and long axis, with phase-sensitive inversion recovery.
The manoeuvre
- Cine: LV and RV volumes, ejection fraction and wall motion by segment.
- LGE short-axis stack, base to apex: bright myocardium — its segment, wall layer (subendocardial, mid-wall, epicardial, transmural) and transmurality (<25%, 25–50%, 51–75%, >75%).
- Match to a territory: LAD, circumflex or RCA distribution → ischaemic; not territorial → non-ischaemic.
- Non-ischaemic patterns: inferolateral epicardial/subepicardial (myocarditis), septal mid-wall stripe (dilated cardiomyopathy), patchy mid-wall at the RV insertion points in a thick wall (hypertrophic cardiomyopathy), diffuse subendocardial with abnormal nulling (amyloid), basal inferolateral mid-wall/epicardial with oedema (sarcoid can be anywhere).
- T2/mapping: oedema that goes with the scar (acute) or not (chronic).
- Thrombus: dark filling defect on early and late gadolinium images, often at an akinetic apex.
What confirms it
- A reproducible bright region on two orthogonal planes, in a pattern that matches either a coronary territory or a named non-ischaemic pattern.
What licenses you to exclude it
- A good-quality study with correct nulling and no LGE makes established scar unlikely; diffuse fibrosis can still be present — native T1 and ECV address it.
The classic misread
- Calling the RV insertion points (small, common) a cardiomyopathy on their own.
- Mis-set inversion time making normal myocardium grey — every segment then looks abnormal.
- Calling amyloid "normal" because the whole myocardium is difficult to null — that difficulty is the sign.