Abstract Body (Do not enter title and authors here): Background:
Electroanatomic mapping (EAM) and cardiac magnetic resonance (CMR) identify substrates using parameters from ischemic cardiomyopathy (ICM). CMR's application to characterize scar in non-ischemic dilated cardiomyopathy (NIDCM) is limited.

Objective: To assess scar burden and patterns in patients with VA and NIDCM or ICM and the concordance between CMR and EAM.

Methods:
We reviewed 27 patients with VA ablations (NIDCM and ICM) and prior CMR. ARVC, HCM and infiltrative diseases were excluded. Scar type and morphology were classified (table 2). Myocardial late gadolinium enhancement (LGE) (> 2SD or 3SD) and T1 values were compared to remote myocardium. EAM defined bipolar scar <1.5 mV and unipolar scar <8.3 mV. Scar locations were assigned using the AHA 17-segment model. Cohen’s Kappa coefficient was used to assess categorical agreement of scar severity between CMR and EAM (table 4).

Results:
Scar distribution in NIDCM was predominant in the mid-myocardial septum and ICM in the sub-endocardium (p <0.001). ICM patients had higher presence of posteromedial papillary muscle scar (p=0.001). The difference in myocardial T1 values at the mid-chamber and mid-septum was not significant, likely influenced by the mid-LV chamber short axis cut. The right ventricle was dilated with reduced function and non-compacted myocardium was thicker in NIDCM. Bipolar scar was predominant in AHA segments 1-3 and 9-10 for NIDCM patients and AHA segments 3-12 in ICM patients. AHA segments 4 and 8-12 had fair correlation between bipolar EAM and CMR. Unipolar mapping showed slight agreement in segments 1-6, 9, 11, 13, 15 and 16.

Conclusions:
Scar is predominant in the mid-myocardial septum in NIDCM compared to subendocardial location in ICM. Comparing CMR, EAM at best correlate fairly to midventricular segments in bipolar map slightly to most segments in unipolar map. Improved mapping tools will likely improve ablation success.