ECGrecover: a Deep Learning Approach for Electrocardiogram Signal Completion

TL;DR

The proposed ECGrecover, a U-Net neural network model trained on a novel composite objective function to address the reconstruction problem, consistently outperformed state-of-the-art methods in standard distortion metrics as well as in preserving critical ECG characteristics, particularly the P, QRS, and T wave coordinates.

Abstract

In this work, we address the challenge of reconstructing the complete 12-lead ECG signal from its incomplete parts. We focus on two main scenarios: (i) reconstructing missing signal segments within an ECG lead and (ii) recovering entire leads from signal in another unique lead. Two emerging clinical applications emphasize the relevance of our work. The first is the increasing need to digitize paper-stored ECGs for utilization in AI-based applications, often limited to digital 12 lead 10s ECGs. The second is the widespread use of wearable devices that record ECGs but typically capture only one or a few leads. In both cases, a non-negligible amount of information is lost or not recorded. Our approach aims to recover this missing signal. We propose ECGrecover, a U-Net neural network model trained on a novel composite objective function to address the reconstruction problem. This function incorporates both spatial and temporal features of the ECG by combining the distance in amplitude and sycnhronization through time between the reconstructed and the real digital signals. We used real-life ECG datasets and through comprehensive assessments compared ECGrecover with three state-of-the-art methods based on generative adversarial networks (EKGAN, Pix2Pix) as well as the CopyPaste strategy. The results demonstrated that ECGrecover consistently outperformed state-of-the-art methods in standard distortion metrics as well as in preserving critical ECG characteristics, particularly the P, QRS, and T wave coordinates.

Figures (14)

• Figure 1: Example of a 12-lead ECG. The machine has recorded a 10-second complete recording for all 12 leads, but in the PDF and paper printed format, only 2.5 seconds are kept for all the leads, in addition to the complete copy of lead II. This is illustrated by the configuration C3 in \ref{['fig:configurations']}).
• Figure 2: Masks applied to ECGs for simulating real-life incomplete ECGs. The green sections represent the primers – the portions of the signal available, while the red sections indicate the parts of the signal that need to be reconstructed. In \ref{['fig:segment_mask']}, the numbers in brackets specify the primers' length.
• Figure 3: Training set and network architecture.
• Figure 4: Lead II signal obtained with different approaches for C$_\text{I}$. Reconstructed leads are in red, the original ones are in black. Top depicts a relatively clean ECG while Bottom, an noisy ECG with baseline wander
• Figure 5: Performance of ECGrecover and the SOTA in their reconstruction capabilities for segment recovery (C1, C3, C5) and lead reconstruction (C$_{\text{I}}$, C$_{\text{II}}$). ECGrecover is significantly better for all metrics (p-value < 2e-16).