Abstract 18816: Predicting an Accurate Intensity Threshold on LGE-MRI for Identification of Left Atrial Fibrosis

Abstract

Background: Late gadolinium enhanced (LGE) MRI identifies left atrial (LA) fibrosis in atrial fibrillation (AF). However, routine measurement of image intensity in “arbitrary units” limits inter-patient and longitudinal comparisons. This study is designed to identify the scar threshold identification on LGE-MRI to improve the uniformity and validity of results. Methods: High-resolution navigator-gated LGE-MRI (1.5 Tesla) was performed in 75 patients (31% female, age 62 ± 8 years) prior to the first (60%), second (24%), or third (16%) ablation for AF. Epicardial and endocardial contours were manually drawn around LA myocardium. The image intensity ratio (IIR) was defined as LA myocardial signal intensity for each of 20 sectors on contiguous axial planes divided by the mean LA blood pool image intensity. Intra-cardiac point-by-point sampled electro-anatomical map (EAM) points were registered and superimposed upon the image. Log transformed bipolar voltage was plotted against IIR and the threshold values corresponding to 0.5mV and 0.1mV were identified. Results: A total of 8153 EAM points and corresponding image sectors were analyzed. In a generalized estimating equations model accounting for within patient clustering, and adjusting for age, LA volume, mass, BMI, gender, CHADSVasc score, AF type, prior ablation, and the delay time between contrast and image acquisition, each unit increase in local IIR was associated with 91.3% decrease in bipolar LA voltage (P<0.001). The IIR was strongly associated with bipolar voltage on EAM in patients with minimal and extensive LA fibrosis (figure top). IIR thresholds of 0.97 and 1.62 corresponding to 0.5 mV (fig-bottom left) and 0.1 mV (fig- bottom right), respectively, were obtained and showed convergence of the 95% CI and a stable threshold with increasing patients. Conclusion: Normalization of LGE-MRI intensity by the mean blood pool intensity results in a stable threshold to identify diffuse fibrosis or dense scar.