Quantification of atrial cardiomyopathy disease severity by electroanatomic voltage mapping and cardiac magnetic resonance imaging

Figures

• 

  Figure 1. Clinical characteristics of patient sub‐groups defined using tertiles of mean fibrosis burden. CAD, coronary artery disease; DM, diabetes mellitus; HF, heart failure; HTN, hypertension; PVD, peripheral vascular disease.

• 

  Figure 2. Example Atrial LGE images and electroanatomic voltage maps. (A) Atrial LGE appearances for each category of qualitative assessment (B) Atrial LGE shells reconstructed from Atrial LGE imaging and visualized using CEMRGapp. (C) Electroanatomic voltage shells reconstructed and visualized using OpenEP. LGE, late gadolinium enhancement.

• 

  Figure 3. Qualitative analysis of Atrial‐LGE and Atrial‐EAVM. (A) Prevalence of atrial dilation increases as visual severity category of Atrial LGE increases. (B) Prevalence of atrial dilatation increases with increasing quartiles of low voltage (C) AF recurrence increases with increasing visual Atrial‐LGE category. (D) AF recurrence is low in the absence of significant low voltage but increases significantly in the presence of any low voltage area (quartile 2). EAVM, electroanatomic voltage mapping; LGE, late gadolinium enhancement.

• 

  Figure 4. Comparison between Atrial‐EAVM and Atrial‐LGE. (A) Linear regression with 95% confidence intervals. R values are Pearson correlation coefficients. (B) Bland–Altman plot showing agreement between the atrial area identified as abnormal by cardiac magnetic resonance imaging (image intensity ratio [IIR] > 0.97 threshold) and the atrial area identified as abnormal by electroanatomic voltage mapping (low voltage area <1.17 mV threshold). EAVM, electroanatomic voltage mapping; LGE, late gadolinium enhancement.

• 

  Figure 5. (A) Median atrial surface area identified as fibrotic by each modality, showing reducing absolute differences between the modalities with increasing overall fibrosis burden. p values represent tests of significance for between‐modality differences in disease severity quantification at each tertile of overall fibrosis burden. (B) Decreasing differences between modalities is shown for each tertile of mean scar (both modalities), with convergence of regression lines at the highest fibrosis burden. (C) Qualitative assessment of Atrial‐LGE and the relationship to quantitative fibrosis area, showing that for both Atrial‐LGE and Atrial‐EAVM, a higher visual grade of MRI LGE is associated with increasing fibrosis burden. EAVM, electroanatomic voltage mapping; LGE, late gadolinium enhancement.

• 

  Figure 6. The relationship between Atrial‐LGE and Atrial‐EAVM for each segment of the atria. The relationship is modest or weak for all segments. The strongest is within the PV antra, the weakest within the septum. Within the roof and posterior wall there appears to be significantly more Atrial‐LGE fibrosis than Atrial‐EAVM fibrosis with a large number of zero values using voltage mapping. EAVM, electroanatomic voltage mapping; LGE, late gadolinium enhancement.

• 

  Figure 7. Time to atrial fibrillation recurrence. (A) Time to recurrence categorizing patients using Atrial‐LGE. (B) Time to recurrence categorizing patients using Atrial‐EAVM. p values from log‐rank tests. Optimal split thresholds are shown for each modality. EAVM, electroanatomic voltage mapping; LGE, late gadolinium enhancement.