Artefacts

Cine bSSFP

ArtefactAppearanceCauseFix
Off-resonance bandingDark bands across the image, worst at the edges of the field and near lung, sternum and devicesB0 inhomogeneity; bSSFP signal nulls where the off-resonance phase accumulated per TR reaches πRe-shim over the heart; adjust centre frequency; shorten TR; spoiled GRE cine (worse blood-myocardium contrast, no banding)
Flow ghostingCopies of bright blood displaced along the phase-encode directionPulsatile flow; incomplete breath-holdSwap phase-encode direction; saturation bands; repeat the breath-hold
Respiratory blurringBlurred endocardial border, doubled diaphragmBreathing during the segmented acquisitionShorter breath-hold (fewer phases, higher acceleration); real-time cine; respiratory-gated free-breathing cine
ArrhythmiaBlurred or stepped edges, dropped frames, blocky appearanceIrregular R-R intervals corrupt the segmented k-spaceArrhythmia rejection window; prospective triggering; real-time cine
Mis-triggeringGhosting, wrong phase order, very short apparent R-RECG trace degraded by the magnetohydrodynamic effect or poor electrode contact; T-wave triggeringRe-position electrodes; vectorcardiographic gating; pulse oximetry gating as a last resort
Wraparound (aliasing)Anatomy from outside the field of view folded onto the imageField of view smaller than the object in the phase-encode directionIncrease FOV; phase oversampling; saturation band over the arms
Parallel imagingFold-over or a band of noise in the centre of the imageAcceleration factor too high for the coil geometry (g-factor); wrong calibrationLower acceleration; re-acquire calibration; change phase-encode direction
Metal susceptibilitySignal void with distortion, sometimes with bright pile-up at the edgeSternal wires, valve prostheses, clips, devicesSpoiled GRE cine; shorter TE; the void is smaller at 1.5T than 3T

Late gadolinium enhancement

ArtefactAppearanceCauseFix
Wrong inversion timeMyocardium grey rather than black (TI too short: signal may invert and mimic enhancement; TI too long: reduced scar contrast)TI not matched to the current myocardial T1, which shortens with time after contrastRepeat the TI scout; phase-sensitive inversion recovery (PSIR), which is insensitive to TI error
Bright blood masking subendocardiumThin subendocardial enhancement invisible against the blood poolBlood and scar both bright on standard LGEDark-blood LGE variants; longer TI; compare with cine wall motion
Motion ghostingCopies of enhancing structures along phase encode; blurred scarBreathing or arrhythmia during the segmented acquisitionSingle-shot LGE with motion correction and averaging (free-breathing); repeat with a shorter breath-hold
Gibbs ringingThin bright or dark line parallel to a sharp edge, most often at the blood-myocardium interfaceTruncation of k-space at the acquisition matrixHigher matrix; check the line moves with the phase-encode direction
Metal susceptibilitySignal void and distortion near the implant; inversion fails locally, so myocardium next to a device appears brightLocal field distortion from the device, wires or valveWideband inversion pulse for ICDs; spoiled GRE readout; shorter TE
WraparoundAnatomy folded onto the heartSmall FOVIncrease FOV or phase oversampling

First-pass perfusion

ArtefactAppearanceCauseFix
Dark rim artefactThin dark subendocardial line, 1 to 2 pixels, appearing as the bolus arrives in the LV cavity and fading over a few framesGibbs truncation and cardiac motion at the sharp blood-myocardium boundary during peak contrastHigher in-plane resolution; see the comparison below; a true defect persists and follows a coronary territory
Respiratory motionSlices drift through the stack, mis-registered framesBreathing during the 40 to 60 heartbeat acquisitionMotion-corrected reconstruction; coach a breath-hold through peak enhancement; free-breathing protocols with registration
Parallel imaging noiseCentral band of noise or fold-over in each frameHigh acceleration in a low-SNR sequenceLower acceleration; alternative reconstruction (kt or compressed sensing)
Off-resonanceDark bands through the myocardium in every frame (bSSFP readout)B0 inhomogeneityShim; spoiled GRE readout, which is standard at 3T
ArrhythmiaMissed or repeated frames; a slice acquired at a different cardiac phaseIrregular R-R with a fixed slice scheduleRead the frame timing; accept if the defect persists across frames; repeat if in doubt

T1, T2 and T2* mapping

ArtefactSequenceAppearanceCause and fix
Mis-registrationT1, T2Bright or dark rim at the endocardium or epicardium on the map; poor fit qualityBreathing or motion between the source images. Motion-corrected fitting; repeat with a better breath-hold; read the fit error map
Heart-rate dependenceMOLLI T1T1 underestimated at high heart rates, and shifted by ectopicsIncomplete recovery between inversions. Heart-rate-adapted schemes (5(3)3), ShMOLLI or SASHA; report the scheme used
Off-resonancebSSFP-based T1 and T2Regional T1 or T2 error, usually inferolateral or near lungB0 inhomogeneity through the bSSFP readout. Shim; check the frequency; exclude the segment
Partial volumeAllElevated values at the blood-myocardium border and in thin wallsContours too close to the border. Erode the ROI; report septal values
SusceptibilityT2*Falsely short T2* near the inferior wall, posterior vein, lung and RV insertionLocal field gradients. Measure the mid-septum only; truncate the late echoes in heavy iron loading; 1.5T
Fat-water swapDixonFat and water images exchanged, regionally or whole-slicePhase unwrapping failure with poor shim. Re-shim and repeat; three-point Dixon

Dark rim artefact versus true subendocardial defect or LGE

FeatureDark rim artefactTrue finding
DistributionCircumferential or at the phase-encode edges, all territoriesFollows a coronary territory
Timing (perfusion)Appears with cavity enhancement, gone within a few framesPersists through and after peak myocardial enhancement
Thickness1 to 2 pixels, uniformWider, transmural extent varies
Phase-encode directionChanges with itDoes not
Wall motionNormalOften abnormal in the same segment
LGENone in the segmentMatching subendocardial LGE in infarction; none in ischaemia without infarction
  1. Ferreira PF, Gatehouse PD, Mohiaddin RH, Firmin DN. Cardiovascular magnetic resonance artefacts. J Cardiovasc Magn Reson 2013;15(1):41. doi:10.1186/1532-429X-15-41
  2. Di Bella EVR, Parker DL, Sinusas AJ. On the dark rim artifact in dynamic contrast-enhanced MRI myocardial perfusion studies. Magn Reson Med 2005;54(5):1295-1299. doi:10.1002/mrm.20666
  3. Kramer CM, Barkhausen J, Bucciarelli-Ducci C, Flamm SD, Kim RJ, Nagel E. Standardized cardiovascular magnetic resonance imaging (CMR) protocols: 2020 update. J Cardiovasc Magn Reson 2020;22(1):17. doi:10.1186/s12968-020-00607-1
  4. Messroghli DR, Moon JC, Ferreira VM, et al. Clinical recommendations for cardiovascular magnetic resonance mapping of T1, T2, T2* and extracellular volume: a consensus statement by SCMR endorsed by EACVI. J Cardiovasc Magn Reson 2017;19(1):75. doi:10.1186/s12968-017-0389-8