A comprehensive evaluation of the left atrium using cardiovascular magnetic resonance

Figures

• 

  Fig. 1. Example PV map of a patient scheduled for AF ablation. RSPV right superior pulmonary vein, MPV middle pulmonary vein (anatomic variant), RIPV right inferior pulmonary vein, LSPV left superior pulmonary vein, LIPV left inferior pulmonary vein

• 

  Fig. 2. LAA thrombus (arrow) identified in a patient with AF undergoing CMR for evaluation of cardiomyopathy. (Left) Ill-defined low signal intensity is seen in the LAA on bSSFP cine. (Right) Due to lack of contrast uptake, the thrombus is of low signal intens on LGE images. Additionally, LGE is seen in the LV apex. bSSFP balanced steady-state free precession, TI inversion time, LGE late gadolinium enhancement, LAA left atrial appendage, CMR cardiovascular magnetic resonance, LV left ventricle

• 

  Fig. 3. (A) 3D LA LGE pulse sequence diagram using stack-of-star k-space sampling; (B) XD-GRASP reconstruction pipeline with self-navigation of respiratory motion; (C) LA LGE images in three orthogonal planes and the corresponding LA fibrosis map using the IIR method with 1.2 cutoff value. 3D three-dimensional, LA left atrium, LGE late gadolinium enhancement, XD-GRASP eXtra-Dimensional Golden-angle RAdial Sparse Parallel, ECG electrocardiogram, TI inversion time, NUFFT non-uniform fast Fourier transform, CS compressed sensing, DE delayed enhancement MRI, cor coronal, sag sagittal, axi axial.

• 

  Fig. 4. (A) Healthy subject. Comparison of bright blood bSSFP (top row, left) with coronal (top row, right) and short-axis (bottom row) dark blood echo-uT1RESS. (B) Patient with apical variant hypertrophic cardiomyopathy, small left ventricular apical aneurysm, and left atrial myxoma (arrow). The mass appears relatively dark with bright blood cine (left) and bright with dark blood unenhanced echo-uT1RESS (middle). After contrast administration, dark blood inversion-prepared echo-uT1RESS (right) shows enhancement of the mass and scar tissue at the left ventricular apex (open arrow). bSSFP balanced steady-state free precession, uT1RESS unbalanced T1 Relaxation-Enhanced Steady-State

• 

  Fig. 5. The appearance of acute ablation lesions in LGE changes with time after GBCA administration. MVO regions (red arrows) significantly decrease with time post GBCA administration. LGE was performed about 1 h after the conclusion of ablation procedure. Axial and sagittal views are shown. LGE late gadolinium enhancement, GBCA gadolinium-based contrast agent, MVO microvascular obstruction

• 

  Fig. 6. (A) Pulse sequence diagram for 3D TWILITE demonstrating a 2-RR interval acquisition with a long inversion time and a respiratory navigator. (B) Longitudinal magnetization evolution for TWILITE showing increased contrast between lesion core (with reduced T1) and normal myocardium after a long inversion time and a low-intensity (dark) blood pool. C. Sample non-contrast enhanced T1-weighted 3D TWILITE images acquired in swine after RF ablation mimicking pulmonary vein isolation. Necrosis from ablation is hyperintense relative to unablated tissues. Yellow: ablations in LA; orange: ablations in right superior pulmonary vein (RSPV); yellow ellipse: ostium of RSPV. (D) Comparison of 3D TWILITE and early and late gadolinium contrast-enhanced after RF ablation in the same swine shown in (C) LGE is complicated by time-varying contrast (signal voids) and enhancement of edema and preexisting scar as well as necrosis. Yellow: anterior ablations; orange: posterior ablations; red arrows: gaps between lesions. 3D three-dimensional, TWILITE T1-weighted long-inversion time, LGE late gadolinium enhancement, ECG electrocardiogram, LA left atrium, LV left ventricle, AO aorta, RV right ventricle.

• 

  Fig. 7. Example of LA 4D hemodynamic signature quantification quantifying 4D flow MRI-derived velocity vector fields in seven dimensions (3D space + 3 velocity directions + time) for a 72-year-old male AF patient vs a 77-year-old female healthy control. (a) LA velocity vector field in the healthy control that an average of healthy hemodynamic signature index (HSI) of 11.8, showing dynamic flow and vortex formation. (b) LA velocity vector field in the AF patient with an elevated HSI of 24, showing reduced dynamics, disturbed flow, and lack of vortex formation. (c) 4D flow MRI-derived hemodynamic signature profiles (analyzing all seven dimensions) of the AF patient (red) compared to the interquartile range of healthy controls (shaded gray), indicating increased lower velocity vector disparities (left skew), reduced flow complexity, and stagnant flow patterns, highlighting impaired flow dynamics in the AF patient. LA left atrium, 4D four-dimensional, MRI magnetic resonance imaging, 3D three-dimensional, AF atrial fibrillation, IQR interquartile range

• 

  Fig. 8. (A) Arrhythmia resolved 5D flow MRI (x, y, z, t, heart rate) utilizes a free running, radial acquisition with 3D velocity encoding. An SI projection is played out every 21 radial lines for inherent cardiac and/or respiratory signal extraction. (B) The atrial fibrillation (AF) cardiac signal, as determined by the SI signal or an external ECG, is subdivided into different heartbeat length bins by an algorithm (e.g. for four bins in this example: RR1, RR2, RR3, and RR4). K-space data are sorted and combined based on these heartbeats and desired temporal resolution for cardiac phases within each heartbeat. (C) Compressed sensing is used to reconstruct each data set (color-coded for each RR bin), yielding a flow data set (x, y, z, vx, vy, vz, t) centered at a specific heart rate. (D) Atrial blood flow can then be pre-processed for each RR bin and the left atrial structure(s) subsequently segmented, allowing for blood flow quantification, such as atrial blood stasis of the left atrium, in this example. 5D five-dimensional, MRI magnetic resonance imaging, 3D three-dimensional, ECG electrocardiogram