Quality assessment of cardiac magnetic resonance myocardial scar imaging prior to ventricular arrhythmia ablation

Figures

• 

  Fig. 1. Illustration of motion and ICD artifacts for conventional LGE MRI techniques used for scar imaging. A1–5 Cardiac gated, breath-hold inversion-recovery (IR) gradient recalled echo (GRE) imaging collected with simulated heart rate 75 bpm. A1 image without motion is sharper than A2 with 2 cm, 3 Hz motion simulating poor breath-holding. Motion causes ghosting artifacts (e.g., red arrows), blurring (e.g., dashed red arrow), and appearance of structures that fall outside the desired image plane (e.g., yellow arrow). A3 placement of an ICD over the stationary object introduces different image artifacts caused by ferromagnetic ICD components. Hyperintensity occurs in areas that fall outside the bandwidth of the IR-pulse (yellow arrow). Distortion of the grid (dashed red arrow) and signal void (red arrow) are seen closer to the ICD, A4 Wider-bandwidth IR reduces hyperintensity artifact (yellow arrow). Higher receiver bandwidth additionally decreases image distortion closer to the ICD. A5 In the presence of an ICD, conspicuous ghosting artifacts are seen during 2 cm simulated respiratory motion (red arrows). B1–3 Single-shot IR balanced steady state-free precession (SSFP) imaging. B1 image…

• 

  Fig. 2. Effects of motion on pre-ablation LGE. A1, 2 Minimal motion artifact during LV short axis imaging. A1 bh-LGE, and A2 and ss-SSFP-LGE. Subtle features like subendocardial scar (solid yellow arrow) can be more difficult to detect on ss-SSFP-LGE images due to lower resolution and higher noise (dashed yellow arrow). B Motion in a patient without ICD. B1 bh-LGE horizontal long axis image where respiratory motion limits distinction of scar from motion artifact (dashed yellow arrows). Short-axis imaging was not performed because of the subjects’ difficulty with breath-holding. B2 Acceptable quality ss-SSFP-LGE short axis imaging could be performed in this subject. Unambiguous septal scarring is highlighted (yellow arrow). C Motion and noise affecting scar interpretation in a patient with an ICD. C.1 Cardiac gated breath-hold image with limited interpretation of possible “patchy” scar (dashed yellow arrow). C.2 ss-SSFP-LGE of the same short axis image slice is less helpful for reducing motion artifacts due to significant banding artifacts caused by ferromagnetic components in the ICD (dashed yellow arrows)

• 

  Fig. 3. Effects of ICDs on pre-ablation LGE. A Acceptable ICD artifact (dashed yellow arrows), that is remote from the LV anterior wall (solid yellow arrow). B, C ICD artifacts limiting scar interpretation (dashed yellow arrows). B The LV anterior wall falls outside the bandwidth of the wide-bandwidth inversion RF pulse resulting in image hyperintensity. C The anterior wall is obscured by signal dropout and image distortion which occurs adjacent to the ICD

• 

  Fig. 4. Contribution of motion artifact, ICD artifact, and contrast quality to imaging studies with limited whole-heart scar interpretation in ICD and non-ICD patients