Background

Periodic repolarization dynamics (PRD) is an electrocardiographic biomarker that quantifies low-frequency (LF) instabilities of repolarization. PRD is a strong predictor of mortality in patients with ischaemic and non-ischaemic cardiomyopathy. Until recently, two methods for calculating PRD have been proposed. The wavelet analysis has been widely tested and quantifies PRD in deg² units by application of continuous wavelet transformation (PRD^wavelet). The phase rectified signal averaging method (PRD^PRSA) is an algebraic method, which quantifies PRD in deg. units. The correlation, as well as a conversion formula between the two methods remain unknown.

Method

The first step for quantifying PRD is to calculate the beat-to-beat change in the direction of repolarization, called dT°. PRD is subsequently quantified by means of either wavelet or PRSA-analysis. We simulated 1.000.000 dT°-signals. For each simulated signal we calculated PRD using the wavelet and PRSA-method. We calculated the ratio between PRD^wavelet and PRD^PRSA for different values of dT° and RR-intervals and applied this ratio in a real-ECG validation cohort of 455 patients after myocardial infarction (MI). We finally calculated the correlation coefficient between real and calculated PRD^wavelet. PRD^wavelet was dichotomized at the established cut-off value of ≥5.75 deg².

Results

The ratio between PRD^wavelet and PRD^PRSA increased with increasing heart-rate and mean dT°-values (p < 0.001 for both). The correlation coefficient between PRD^wavelet and PRD^PRSA in the validation cohort was 0.908 (95% CI 0.891–0.923), which significantly (p < 0.001) improved to 0.945 (95% CI 0.935–0.955) after applying the formula considering the ratio between PRD^wavelet and PRD^PRSA obtained from the simulation cohort. The calculated PRD^wavelet correctly classified 98% of the patients as low-risk and 87% of the patients as high-risk and correctly identified 97% of high-risk patients, who died within the follow-up period.

Conclusion

This is the first analytical investigation of the different methods used to calculate PRD using simulated and clinical data. In this article we propose a novel algorithm for converting PRD^PRSA to the widely validated PRD^wavelet, which could unify the calculation methods and cut-offs for PRD.

中文翻译：

---

周期性复极动力学：量化复极低频振荡的不同方法

背景

周期性复极动力学 (PRD) 是一种心电图生物标志物，可量化复极的低频 (LF) 不稳定性。PRD 是缺血性和非缺血性心肌病患者死亡率的有力预测因子。直到最近，才提出了两种计算 PRD 的方法。小波分析已得到广泛测试，并通过应用连续小波变换 (PRD^{小波) 以 deg 2}为单位量化 PRD 。相位整流信号平均法 (PRD ^PRSA ) 是一种代数方法，以度为单位量化 PRD。单位。两种方法之间的相关性以及转换公式仍然未知。

方法

量化 PRD 的第一步是计算复极化方向上的逐搏变化，称为dT°。随后通过小波或 PRSA 分析对 PRD 进行量化。我们模拟了 1.000.000 dT°-信号。对于每个模拟信号，我们使用小波和 PRSA 方法计算 PRD。我们计算了不同dT°和 RR 间隔值下 PRD^小波和 PRD ^PRSA之间的比率，并将该比率应用于 455 名心肌梗塞 (MI) 后患者的真实心电图验证队列中。^{我们最终计算了实际和计算的 PRD小波}之间的相关系数。PRD^{小波在设定的≥5.75 deg 2}的截止值处二分。

结果

^PRD小波和 PRD ^PRSA之间的比率随着心率和平均dT°值的增加而增加（ 两者的p < 0.001）。验证队列中PRD^小波和 PRD ^PRSA之间的相关系数为 0.908 (95% CI 0.891–0.923)， 在应用考虑以下比率的公式后，显着 ( p < 0.001) 改善至 0.945 (95% CI 0.935–0.955)从模拟队列中获得的PRD^小波和 PRD ^{PRSA 。}计算出的 PRD^小波正确地将 98% 的患者分类为低风险，87% 的患者分类为高风险，并正确识别了 97% 的高风险患者，这些患者在随访期内死亡。

结论

这是使用模拟和临床数据计算 PRD 的不同方法的首次分析研究。在本文中，我们提出了一种将 PRD ^PRSA转换为广泛验证的 PRD^小波的新算法，该算法可以统一 PRD 的计算方法和截止值。