Search

Background: Left atrial (LA) late gadolinium enhancement (LGE) magnetic resonance imaging (MRI) has varying protocols, which may limit interstudy comparisons. This study evaluates whether changes in image acquisition and reconstruction parameters affect quantification of LA LGE MRI. Methods: A total of 200 LGE MRI scans were prepared using various techniques from 5 subjects. 3D respiratory and ECG-gated LGE images were acquired at 15, 20, 25, and 30 minutes after contrast. Multiplanar reconstructions (MPR) and maximum intensity projections (MIP) were created from 3D data at various slice thicknesses. LA LGE was quantified with image intensity ratio (IIR, atrial wall signal intensity divided by blood pool). Mean IIR was calculated for the entire LA as a metric of LGE burden. Additionally, in 42 consecutive preablation patients, local IIR values from 3.5-mm MPR and MIP reconstructions were separately registered to LA electroanatomic voltage maps (EAM) obtained prior to ablation procedure. Results: Multitime point LGE-MRI showed a rise in baseline mean IIR of 0.5% ± 0.1% per minute (P <.001) in MPR and a rise of 0.6% ± 0.1% (P <.001) for MIP using regression analysis (Figure A). The MPR- and MIP-based mean IIRs diverged, showing change of –0.3% for MPR and +1.1% for MIP for 1-mm increase in slice thickness (P <.001 for both; Figure B). A total of 4428 EAM points were registered to local LA wall IIRs from both MPR and MIP techniques. An increase of a single unit of IIR was related to drop in 1.20 mV with MPR and 1.25 mV with MIP (P <.001 for both). Conclusions: Variations in timing of acquisition, image reconstruction methods, and slice thickness result in changes in LA LGE quantity. Display Omitted [Copyright &y& Elsevier]