Increasing averaging beats improves the test accuracy on Holter‐based late potentials in patients with myocardial infarction

Figures

• 

  FIGURE 1. In the MI‐VT group, the positive rate of LP significantly increased at 800 averaging beats (p = .008 vs. 250 averaging beats) at night‐time. However, there were no significant changes in the LP positive rate in the MI non‐VT and healthy groups at night‐time (a). During daytime, the positive rate of LP significantly increased by 800 averaging beats (p = .001 vs. 250 beats averaging). No significant changes were observed in the LP positive rate in the MI non‐VT and normal control groups at night‐time (b). LP, late potentials; MI‐VT group, post‐myocardial infarction ventricular tachycardia group; MI non‐VT group, post‐myocardial infarction non‐ventricular tachycardia group.

• 

  FIGURE 2. ROC curves for the averaging beats (250, 500, and 800 beats). At night‐time, the AUC was the highest (AUC = 0.828) for averaging beats 800 (a), and the highest (AUC = 0.845) for averaging beats 800 in daytime (b).

• 

  FIGURE 3. Three representative cases of MI‐VT, MI non‐VT, and normal control. In total, 250 beats averaging of MI‐VT resulted in a high base noise level (0.883 μV) and a negative LP result. Subsequent 500 and 800 beats averaging of MI‐VT resulted in reduction of the noise level (0.478 and 0.387 μV for MI non‐VT and normal control participants, respectively). After averaging 500 and 800 beats, the unclear off set (terminal portion) of LP by the noise level became substantial. The LP was positive (a). Conversely, in the MI non‐VT case and healthy case, the LP did not become apparent even if the noise of the base was decreased by averaging 500 and 800 beats, and the LP result was negative in all settings (averaging 250, 500, and 800 beats) (b, c). MI non‐VT group, post‐myocardial infarction non‐ventricular tachycardia group; MI‐VT group, post‐myocardial infarction ventricular tachycardia group.